Abstract
Aims
Ketone bodies (KB) are an important alternative metabolic fuel source for the myocardium. Experimental and human investigations suggest that KB may have protective effects in patients with heart failure. This study aimed to examine the association between KB and cardiovascular outcomes and mortality in an ethnically diverse population free from cardiovascular disease (CVD).
Methods and results
This analysis included 6796 participants (mean age 62 ± 10 years, 53% women) from the Multi-Ethnic Study of Atherosclerosis. Total KB was measured by nuclear magnetic resonance spectroscopy. Multivariable-adjusted Cox proportional hazard models were used to examine the association of total KB with cardiovascular outcomes. At a mean follow-up of 13.6 years, after adjusting for traditional CVD risk factors, increasing total KB was associated with a higher rate of hard CVD, defined as a composite of myocardial infarction, resuscitated cardiac arrest, stroke, and cardiovascular death, and all CVD (additionally included adjudicated angina) [hazard ratio, HR (95% confidence interval, CI): 1.54 (1.12–2.12) and 1.37 (1.04–1.80) per 10-fold increase in total KB, respectively]. Participants also experienced an 87% (95% CI: 1.17–2.97) increased rate of CVD mortality and an 81% (1.45–2.23) increased rate of all-cause mortality per 10-fold increase in total KB. Moreover, a higher rate of incident heart failure was observed with increasing total KB [1.68 (1.07–2.65), per 10-fold increase in total KB].
Conclusion
The study found that elevated endogenous KB in a healthy community-based population is associated with a higher rate of CVD and mortality. Ketone bodies could serve as a potential biomarker for cardiovascular risk assessment.
Keywords: Ketone bodies, Cardiovascular disease, Mortality, Heart failure
Structured Graphical Abstract
Structured Graphical Abstract.
Association between total ketone bodies and cardiovascular events and mortality. CVD, cardiovascular disease; HF, heart failure; KB, ketone bodies; MI, myocardial infarction; NMR, nuclear magnetic resonance.
See the editorial comment for this article ‘Circulating ketone bodies as signals for cardiovascular disease and mortality’, by S.N. Voorrips and B.D. Westenbrink, https://doi.org/10.1093/eurheartj/ehad176.
Introduction
Ketone bodies (KB) [acetoacetate, acetone, β-hydroxybutyrate (β-OHB)] are an important alternative metabolic fuel source for the myocardium.1 Ketone bodies are primarily generated in the liver from fatty acids and delivered to extrahepatic tissues. Ketone bodies metabolism increases in various physiological states such as pregnancy, exercise, and fasting.2 Ketone bodies metabolism has also been suggested to be potentially implicated in numerous human disease states relevant to cardiovascular disease (CVD), including obesity, diabetes, atherosclerosis, and heart failure (HF).3,4 Interestingly, KB are a more efficient energy source than glucose or free fatty acids because they provide more energy per metabolized carbon than glucose and require less oxygen than free fatty acids for ATP generation.5 Ketone bodies also reduce oxidative stress and mitochondrial uncoupling. Moreover, defects in either the synthetic or oxidative arms of KB metabolism may influence cardiovascular energetic machinery.6,7
Evidence suggests that the metabolic shift towards increased KB utilization may increase myocardial blood flow.8 According to data from both experimental and human investigations, KB, β-OHB in particular, appear to have protective effects in patients with CVD.8–10 Furthermore, the profound cardio-protective effects of the sodium-glucose cotransporter 2 inhibitor (SGLT2i) class are hypothesized to be due, in part, to a metabolic switch in favor of KB.10–12 On the other hand, previous studies have demonstrated the association of elevated levels of KB with disease severity and poor prognosis in patients with HF and arrhythmogenic right ventricular cardiomyopathy13–15 and even in patients presenting with ST-segment elevation myocardial infarction (MI) where increased circulating KB at 24 h were independently associated with larger myocardial infarct size and lower left ventricular ejection fraction.16
We aimed to explore the long-term association between circulating KB and cardiovascular outcomes and all-cause mortality in an ethnically diverse population free from CVD at baseline.
Methods
Study population
The Multi-Ethnic Study of Atherosclerosis (MESA) is an ongoing prospective study of 6814 subjects of four racial/ethnic groups (Non-Hispanic White, Black/African American, Hispanic, and Chinese American) living in the USA.17 Recruitment occurred between 2000 and 2002 (Exam 1) from six US field centers (New York, New York; Baltimore, Maryland; Chicago, Illinois; Los Angeles, California; St. Paul, Minnesota; and Forsyth County, North Carolina). Further follow-up exams were conducted from 2002 to 2004 (Exam 2), 2004 to 2005 (Exam 3), 2005 to 2007 (Exam 4), 2010 to 2012 (Exam 5), and 2016 to 2018 (Exam 6). MESA was approved by institutional committees in each field center, and all subjects provided written informed consent upon entry. Further details regarding the MESA design can be found elsewhere.17 All MESA participants with available baseline data for KB were considered for inclusion in the present study. Our final analysis included 6796 participants, excluding those with missing values for KB (n = 18).
Cardiovascular risk factors
All risk factors were measured at the baseline examination. Age, sex, race/ethnicity, smoking status, highest level of educational attainment, medical history, and medication use data were obtained using questionnaires.17 We categorized smoking status as never, former and current smoker. Current smoking was defined as having smoked a cigarette in the past 30 days. A fasting glucose ≥126 mg/dL or antihyperglycemic medication use was used to define diabetes mellitus. Resting blood pressure (BP) was measured three times in the seated position using a Dinamap model Pro 100 automated oscillometric sphygmomanometer (Critikon, Tampa, FL, USA), and the average of the second and third readings was recorded. Hypertension was defined as a BP ≥140/90 mmHg or a history of hypertension and the use of BP medications. Height was measured using a stadiometer, and weight was measured using a standard balance-beam scale or a digital weighing scale. Body mass index (BMI) was calculated using weight (in kilograms) divided by height (in meters) squared. The MESA Typical Week Physical Activity Survey (TWPAS), completed during the baseline examination, identifies the amount of time spent in and the frequency of various physical activities during a typical week in the previous month. Minutes of activity were summed for each discrete activity type and multiplied by the metabolic equivalent (MET) level. In our study, we used total intentional exercise. At the baseline examination, participants completed a 120-item food-frequency questionnaire (FFQ) quantifying usual intake over the past year.18 We used total energy intake in our analysis. Total and high-density lipoprotein cholesterol (HDL-C) and triglyceride levels were measured from blood samples taken after a 12-h fast. Low-density lipoprotein cholesterol (LDL-C) level was estimated according to the Friedewald equation. The estimated glomerular filtration rate (eGFR) was calculated using the 2009 Chronic Kidney Disease Epidemiology Collaboration equation.19 N-terminal pro B-type natriuretic peptide (NT-proBNP) and high-sensitivity cardiac troponin (hs-cTnT) levels were measured at the University of Maryland using the Cobas e601 automated analyzer (Roche Diagnostics).20
Measurement of ketone bodies
Ethylenediaminetetraacetic acid (EDTA) plasma samples that were collected after an overnight fast at baseline were used to measure β-OHB, acetoacetate, and acetone using a nuclear magnetic resonance (NMR) spectroscopy assay.21 EDTA plasma samples from fasting subjects at baseline (2000–02) were stored frozen at <−70° C until NMR testing in 2003. Samples did not undergo any freeze-thaw cycles. The three KB give rise to resolved NMR signals that serve as the basis of their quantification. NMR and mass spectrometry-measured values are highly correlated: R2 = 0.996, 0.994, and 0.994 for β-OHB, acetoacetate, and acetone, respectively.21 Plasma levels of the three KB are conventionally summed as plasma total KB in epidemiologic analyses. The details on the stability of KB at different temperatures and coefficients of variation have been reported previously.21
Cardiovascular disease events
Cardiovascular disease events were independently adjudicated by two physician reviewers and recorded during a mean follow-up of 13.6 years. We used time from baseline to the incidence of particular cardiovascular events for all outcome variables. Events analyzed in this study included all CVD, hard CVD, HF, MI, stroke, CVD mortality, and all-cause mortality. Hard CVD events were defined as a composite of MI, resuscitated cardiac arrest, stroke, and cardiovascular death [secondary to stroke, coronary heart disease (CHD), other atherosclerotic death, or other CVD death]. All CVD events additionally included adjudicated angina.17 A panel reviewed medical records and adjudicated HF events according to standardized criteria. The present study included probable or definite HF events. Probable HF required HF symptoms, a physician diagnosis of HF, and treatment. Definite HF required at least one additional finding, including abnormalities on chest X-ray, echocardiography, or ventriculography. Stroke events included both nonfatal and fatal strokes but not transient ischemic attacks. Identification of cardiovascular events was based on participant self-report via telephone follow-up visits, which were performed at 9–12-month intervals.22 A trained interviewer administered these telephone interviews in a standardized manner to determine whether new CVD diagnoses, hospital admissions, or deaths had occurred. Further events were identified by personal participant notification, the National Death Index, public notices, and MESA clinic visits.22 Medical records and death certificates were requested for all cases. When relevant, interviewers contacted the next of kin in order to obtain copies of death certificates. Finally, two blinded members of the MESA mortality and morbidity review committee performed an independent classification of cardiovascular events and assigned incidence dates. In case of disagreement, adjudication was performed by the full MESA mortality and morbidity review committee.22
Statistical analyses
The baseline characteristics of study participants were described as count (proportion) for categorical variables, and continuous variables were presented as mean (standard deviation) or median (interquartile range) depending on the variable distribution. These characteristics were compared across total KB quartiles at Exam 1 using analysis of variance (ANOVA) for normally distributed data, the Kruskal–Wallis test for skewed distributed data, and the χ2 test for categorical variables. We also reported Spearman correlation coefficients (ρ) of continuous total KB and each of the continuous baseline characteristics. Total KB, β-OHB, and acetoacetate were all log-base 10 transformed to yield an approximately normal distribution and to allow for data interpretation per 10-fold increase of total and individual KB. As total KB, NT-proBNP, and hs-cTnT have skewed distribution; we used the Spearman correlation coefficient to examine the correlation between the logarithmic value of total KB and NT-proBNP and hs-cTnT to determine whether KB provided orthogonal information to these other established CVD biomarkers.
We used the cumulative incidence function (CIF) to estimate the incidence of different outcomes over time in the presence of competing risks.23 The competing risk event was non-CVD death in examining the association between total KB and CVD death and all-cause death in examining the association with other outcomes. Time of censoring was defined as the last follow-up completed prior to 31 December 2017. In addition, we used a restricted cubic spline model adjusted for the variables in Model 3 to characterize the non-linear association between baseline log-total KB (continuous) and each CVD outcome. A test of linearity on the relationship between total KB and each of the CVD outcomes was also done. The association of total KB with cardiovascular events was evaluated with Cox proportional hazards regression models. For each outcome of interest, three Cox models were used to study the relationship between total plasma KB at Exam 1 to the minimum of observed outcome time or time of censoring. The three Cox models included: (i) Model 1: an unadjusted model, (ii) Model 2 adjusted for age, sex, race/ethnicity, and education level (≤high school or above), and (iii) Model 3, additionally adjusted for BMI, diabetes, smoking, systolic BP, BP medications, total cholesterol, HDL-C, lipid-lowering therapy, eGFR, physical activity, total calorie intake, and NT-proBNP.
Additionally, interactions for sex, diabetes, BMI, and hypertension were tested at P < 0.10. A two-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, USA).
Results
Baseline characteristics
A total of 6796 participants (mean age 62.2 ± 10.2, 53% women) were included in this analysis. Overall, 2617 (38.5%) participants were White, 1881 (27.7%) were Black, 1494 (22.0%) were Hispanic, and 804 were Chinese American (11.8%). Of note, most participants were overweight at enrollment (mean BMI 28.3 ± 5.5 kg/m2). Generally, older and female participants had higher KB levels. In addition, participants in the higher quartiles of total KB levels were more likely to have lower LDL-C levels, higher HDL-C levels, lower triglycerides, and lower total cholesterol but had a higher prevalence of diabetes and hypertension. Table 1, Supplementary material online, Table S1 and Figures S1–S3 depict the baseline characteristics of the participants in this analysis. There were 7 missing values for NT-proBNP and 15 for hs-cTnT. Poor correlations were observed between KB and NT-proBNP and hs-cTnT levels. The Spearman correlation coefficient for the total KB and NT-proBNP was 0.15 (see Supplementary material online, Figure S4) and for the total KB and hs-cTnT was 0.09 (see Supplementary material online, Figure S5).
Table 1.
Baseline characteristics by total ketone bodies quartiles
Q1 (<159 µmol/L) (n = 1719) | Q2 (159–213 µmol/L) (n = 1671) | Q3 (213.1–316 µmol/L) (n = 1709) | Q4 (>316 µmol/L) (n = 1697) | ρ | P-value | |
---|---|---|---|---|---|---|
Age (years), mean ± SD | 59.4 ± 9.78 | 61.4 ± 10.06 | 63.0 ± 9.96 | 64.9 ± 10.30 | 0.205 | <0.001 |
Gender, no. (%) | NA | <0.001 | ||||
ȃFemale | 806 (46.9) | 806 (48.2) | 976 (57.1) | 1000 (58.9) | ||
ȃMale | 913 (53.1) | 865 (51.8) | 733 (42.9) | 697 (41.1) | ||
Race/ethnicity, no. (%) | NA | <0.001 | ||||
ȃNon-Hispanic White | 682 (39.7) | 611 (36.6) | 634 (37.1) | 690 (40.7) | ||
ȃChinese-American | 241 (14.0) | 218 (13.0) | 190 (11.1) | 155 (9.1) | ||
ȃAfrican-American | 432 (25.1) | 451 (27.0) | 495 (29.0) | 503 (29.6) | ||
ȃHispanic | 364 (21.2) | 391 (23.4) | 390 (22.8) | 349 (20.6) | ||
Body mass index, mean ± SD, kg/m² | 27.7 ± 4.85 | 28.7 ± 5.22 | 28.8 ± 5.88 | 28.1 ± 5.82 | 0.010 | <0.001 |
Education, no. (%) | NA | 0.305 | ||||
ȃBelow high school education | 297 (17.3) | 305 (18.3) | 337 (19.7) | 305 (18.0) | ||
ȃAbove high school education | 1422 (82.7) | 1366 (81.7) | 1372 (80.3) | 1392 (82.0) | ||
Systolic blood pressure (mmHg), mean ± SD | 123.0 ± 20.41 | 125.9 ± 20.51 | 128.0 ± 21.88 | 129.4 ± 22.52 | 0.113 | <0.001 |
Taking hypertension medication, no. (%) | NA | <0.001 | ||||
ȃNo | 1188 (69.1) | 1082 (64.8) | 1010 (59.1) | 990 (58.3) | ||
ȃYes | 531 (30.9) | 589 (35.2) | 699 (40.9) | 707 (41.7) | ||
Hypertension, no. (%) | NA | <0.001 | ||||
ȃNo | 1073 (62.4) | 955 (57.2) | 879 (51.4) | 841 (49.6) | ||
ȃYes | 646 (37.6) | 716 (42.8) | 830 (48.6) | 856 (50.4) | ||
Diabetes, no. (%) | NA | <0.001 | ||||
ȃNo diabetes | 1610 (93.7) | 1475 (88.3) | 1444 (84.5) | 1405 (82.8) | ||
ȃDiabetes | 109 (6.3) | 196 (11.7) | 265 (15.5) | 292 (17.2) | ||
Fasting glucose (mg/dL), mean ± SD | 91.9 ± 20.55 | 96.5 ± 24.92 | 99.8 ± 29.93 | 101.4 ± 40.98 | 0.073 | <0.001 |
LDL cholesterol (mg/dL), mean ± SD | 118.7 ± 30.27 | 118.1 ± 31.20 | 116.7 ± 31.70 | 115.2 ± 32.43 | −0.047 | 0.002 |
HDL cholesterol (mg/dL), mean ± SD | 49.4 ± 14.23 | 49.0 ± 13.63 | 50.8 ± 14.48 | 54.6 ± 16.20 | 0.130 | <0.001 |
Total cholesterol (mg/dL), mean ± SD | 195.5 ± 37.00 | 194.9 ± 34.37 | 193.5 ± 35.75 | 192.7 ± 35.61 | −0.028 | 0.143 |
Triglycerides (mg/dL), median (IQR) | 115.0 (81.0–165.0) | 122.0 (86.0–172.0) | 113.0 (80.0–161.0) | 96.0 (68.0–142.0) | −0.124 | <0.001 |
Taking lipid lowering medication, no. (%) | NA | 0.089 | ||||
ȃNo | 1466 (85.3) | 1375 (82.3) | 1418 (83.0) | 1430 (84.3) | ||
ȃYes | 253 (14.7) | 296 (17.7) | 291 (17.0) | 267 (15.7) | ||
Cigarette smoking status, no. (%) | NA | 0.931 | ||||
ȃNever | 873 (50.8) | 819 (49.0) | 872 (51.0) | 855 (50.4) | ||
ȃFormer | 622 (36.2) | 624 (37.3) | 619 (36.2) | 626 (36.9) | ||
ȃCurrent | 224 (13.0) | 228 (13.6) | 218 (12.8) | 216 (12.7) | ||
Total intentional exercise (MET/week), mean ± SD | 1558.5 ± 2358.0 | 1629.3 ± 2324.8 | 1498.2 ± 2394.0 | 1529.9 ± 2282.5 | −0.018 | 0.138 |
Total energy intake (kcal), mean ± SD | 1655.3 ± 889.50 | 1654.3 ± 887.71 | 1632.5 ± 882.03 | 1549.6 ± 817.28 | −0.039 | 0.001 |
Estimated glomerular filtration rate (mL/min/1.73 m2), mean ± SD | 74.6 ± 17.54 | 74.3 ± 15.05 | 74.7 ± 16.21 | 74.1 ± 17.02 | −0.013 | 0.239 |
HDL, high-density lipoprotein; IQR, interquartile range; LDL, low-density lipoprotein; MET, metabolic equivalent; SD, standard deviation.
ρ, Spearman's rank correlation coefficient.
Association of plasma ketone bodies with cardiovascular disease events
The 6796 participants included in this analysis were followed for cardiovascular events for a median duration of 15.75 (0.02–17.5) years. The cumulative incidence of HF, stroke (Figure 1), all CVD, hard CVD, and CVD-related death (Figure 2) was higher among individuals in the higher total KB quartiles.
Figure 1.
Cumulative incidence curves of (A) heart failure, (B) stroke, (C) myocardial infarction stratified by ‘Total ketone body’ quartiles (after adjusting for competing risk) CVD, cardiovascular disease; HF, heart failure; MI, myocardial infarction.
Figure 2.
Cumulative incidence curves of (A) all cardiovascular disease, (B) hard cardiovascular diseasecardiovascular disease, (C) cardiovascular disease-related death stratified by ‘Total ketone body’ quartiles (after adjusting for competing risk) CVD, cardiovascular disease.
The restricted cubic splines analysis, evaluating the association of total KB across continuous distribution and different outcomes, demonstrated a significantly higher risk of all CVD, hard CVD, CVD death, all-cause death, and HF at higher levels of total KB. The test of non-linearity was not statistically significant for all CVD (P = 0.26), hard CVD (P = 0.26), HF (P = 0.07), stroke (P = 0.21), CVD death (P = 0.09), and all-cause death (P = 0.06). The test of non-linear association was statistically significant for MI (P < 0.01), suggesting a non-linear association between total KB and MI. (Figures 3 and 4).
Figure 3.
Restricted cubic spline of the association between total ketone bodies and (A) all cardiovascular disease, (B) hard cardiovascular disease, (C) cardiovascular disease-related death and (D) all-cause death cardiovascular disease, cardiovascular disease model adjusted for age, sex, race/ethnicity, education level, body mass index, diabetes, smoking, systolic blood pressure, blood pressure medications, total cholesterol, high-density lipoprotein, lipid-lowering therapy, estimated glomerular filtration rate, physical activity, total calorie intake, and NT-proBNP. Total ketone body concentration was restricted to a maximal value of 2000 μmol/L after excluding five outliers for better visualization. The three knots were placed at 5th, 50th, and 95th percentiles.
Figure 4.
Restricted cubic spline of the association between total ketone bodies and (A) heart failure, (B) stroke and (C) myocardial infarction model adjusted for age, sex, race/ethnicity, education level, body mass index, diabetes, smoking, systolic blood pressure, blood pressure medications, total cholesterol, high-density lipoprotein, lipid-lowering therapy, estimated glomerular filtration rate, physical activity, total calorie intake, and NT-proBNP. Total ketone body concentration was restricted to a maximal value of 2000 μmol/L after excluding five outliers for better visualization. The three knots were placed at 5th, 50th, and 95th percentiles.
Every 10-fold increase in total plasma KB at baseline was associated with a 2.05-fold increased hazard for all CVD events and a 2.34-fold increased hazard for hard CVD events (Table 2, Model 1). This association remained significant after adjusting for age, sex, race/ethnicity, and education (Table 2, Model 2), as well as after further adjusting for BMI, diabetes, smoking, systolic BP, BP medications, total cholesterol, HDL-C, lipid-lowering therapy, eGFR, physical activity, total calorie intake, and NT-proBNP [hazard ratio (HR) 1.37 (95% confidence interval, CI 1.04–1.80) and 1.54 (95% CI 1.12–2.12, per 10-fold increase in total KB) respectively] (Table 2, Model 3). We also observed a statistically significant association between higher total KB level and incident HF [HR 1.68 (95% CI 1.07–2.65, per 10-fold increase in total KB)], in the fully adjusted model (Table 2, Model 3). In contrast to HF, hard CVD, and all CVD, no significant associations were observed between total KB and incident MI and stroke.
Table 2.
Association between total ketone bodies and incident events
Total number of events | Total KB per 10-fold increase | ||
---|---|---|---|
HR (95% CI) | P-value | ||
All CVD | 1042 | ||
Model 1 | 2.05 (1.60–2.63) | <0.001 | |
Model 2 | 1.40 (1.08–1.82) | 0.012 | |
Model 3 | 1.37 (1.04–1.80) | 0.025 | |
Hard CVD | 757 | ||
Model 1 | 2.34 (1.75–3.12) | <0.001 | |
Model 2 | 1.59 (1.17–2.16) | 0.003 | |
Model 3 | 1.54 (1.12–2.12) | 0.008 | |
HF | 385 | ||
Model 1 | 3.10 (2.09–4.59) | <0.001 | |
Model 2 | 1.93 (1.27–2.93) | 0.002 | |
Model 3 | 1.68 (1.07–2.65) | 0.025 | |
MI | 336 | ||
Model 1 | 1.53 (0.98–2.40) | 0.064 | |
Model 2 | 1.16 (0.73–1.86) | 0.532 | |
Model 3 | 1.06 (0.64–1.75) | 0.818 | |
Stroke | 307 | ||
Model 1 | 2.66 (1.71–4.16) | <0.001 | |
Model 2 | 1.68 (1.04–2.70) | 0.034 | |
Model 3 | 1.56 (0.94–2.61) | 0.087 | |
All-cause death | 1548 | ||
Model 1 | 3.43 (2.82–4.16) | <0.001 | |
Model 2 | 1.89 (1.53–2.32) | <0.001 | |
Model 3 | 1.81 (1.45–2.23) | <0.001 | |
CVD death | 359 | ||
Model 1 | 3.67 (2.47–5.46) | <0.001 | |
Model 2 | 1.96 (1.28–3.01) | 0.002 | |
Model 3 | 1.87 (1.17–2.97) | 0.008 |
CVD, cardiovascular disease; HF, heart failure; HR, hazard ratio; KB, ketone bodies; MI, myocardial infarction; NT-proBNP, N-terminal pro B-type natriuretic peptide; SD, standard deviation.
Model 1: Unadjusted.
Model 2: Adjusted for age, sex, race/ethnicity, education level (≤high school or above).
Model 3: Model 2 + body mass index, diabetes, smoking, systolic blood pressure, blood pressure medications, total cholesterol, high-density lipoprotein cholesterol, lipid-lowering therapy, estimated glomerular filtration rate, physical activity, total calorie intake, and NT-proBNP.
Bold indicates P-value < 0.05.
The associations between total KB at baseline with CVD-related death and all-cause death were next explored. In fully adjusted analysis, MESA participants with higher total KB had significantly higher hazards of CVD-related death [HR 1.87 (95% CI 1.17–2.97)] and all-cause mortality [HR 1.81 (95% CI 1.45–2.23)] per 10-fold increase in total KB (Table 2).
Sensitivity and subgroup analysis
We examined the association between individual KB, including β-OHB and acetoacetate, which are more easily available tests in clinical laboratories and incident events (Table 3). Compared with total KB, the association between individual KB and incident events remained similar, with increasing β-OHB and acetoacetate associated with all-CVD, hard-CVD, HF, all-cause death, and CVD-death but not with MI and stroke. There was no significant interaction of sex, diabetes, BMI, or hypertension on the association of total KB levels with all cardiovascular outcomes (P > 0.10 for all).
Table 3.
Association between β-hydroxybutyrate, acetoacetate and incident events
β-OHB per 10-fold increase | Acetoacetate per 10-fold increase | |||
---|---|---|---|---|
HR (95% CI) | P-value | HR (95% CI) | P-value | |
All CVD | ||||
Model 1 | 1.88 (1.48–2.40) | <0.001 | 1.86 (1.51–2.30) | <0.001 |
Model 2 | 1.32 (1.02–1.70) | 0.034 | 1.42 (1.14–1.75) | 0.001 |
Model 3 | 1.33 (1.02–1.73) | 0.034 | 1.41 (1.14–1.76) | 0.002 |
Hard CVD | ||||
Model 1 | 2.14 (1.61–2.84) | <0.001 | 2.05 (1.60–2.63) | <0.001 |
Model 2 | 1.49 (1.10–2.00) | 0.009 | 1.56 (1.22–2.01) | <0.001 |
Model 3 | 1.42 (1.04–1.94) | 0.026 | 1.52 (1.18–1.97) | 0.001 |
HF | ||||
Model 1 | 3.00 (2.05–4.41) | <0.001 | 2.42 (1.72–3.42) | <0.001 |
Model 2 | 1.90 (1.26–2.86) | 0.002 | 1.76 (1.24–2.48) | 0.002 |
Model 3 | 1.64 (1.05–2.55) | 0.030 | 1.56 (1.07–2.25) | 0.020 |
MI | ||||
Model 1 | 1.30 (0.86–2.06) | 0.200 | 1.66 (1.14–2.41) | 0.007 |
Model 2 | 1.04 (0.66–1.64) | 0.856 | 1.35 (0.93–1.97) | 0.118 |
Model 3 | 0.95 (0.59–1.53) | 0.832 | 1.26 (0.85–1.88) | 0.255 |
Stroke | ||||
Model 1 | 2.57 (1.67–3.97) | <0.001 | 2.04 (1.38–3.01) | <0.001 |
Model 2 | 1.62 (1.02–2.58) | 0.041 | 1.51 (1.02–2.24) | 0.040 |
Model 3 | 1.51 (0.92–2.48) | 0.107 | 1.48 (0.98–2.24) | 0.064 |
All-cause death | ||||
Model 1 | 3.34 (2.77–4.04) | <0.001 | 2.35 (1.98–2.79) | <0.001 |
Model 2 | 1.85 (1.51–2.27) | <0.001 | 1.54 (1.30–1.84) | <0.001 |
Model 3 | 1.77 (1.43–2.20) | <0.001 | 1.49 (1.24–1.79) | <0.001 |
CVD death | ||||
Model 1 | 3.48 (2.36–5.14) | <0.001 | 2.84 (1.99–4.05) | <0.001 |
Model 2 | 1.86 (1.22–2.83) | 0.004 | 1.83 (1.28–2.62) | 0.001 |
Model 3 | 1.77 (1.12–2.79) | 0.014 | 1.71 (1.17–2.49) | 0.006 |
β-OHB, beta-hydroxybutyrate; CVD, cardiovascular disease; HF, heart failure; HR, hazard ratio; MI, myocardial infarction; NT-proBNP, N-terminal pro B-type natriuretic peptide; SD, standard deviation.
Model 1: Unadjusted.
Model 2: Adjusted for age, sex, race/ethnicity, education level (≤high school or above).
Model 3: Model 2 + body mass index, diabetes, smoking, systolic blood pressure, blood pressure medications, total cholesterol, high-density lipoprotein cholesterol, lipid-lowering therapy, estimated glomerular filtration rate, physical activity, total calorie intake, and NT-proBNP.
Bold indicates P-value < 0.05.
Discussion
Experimental and clinical research suggests that circulating ketones may be beneficial for patients with CVD; however, the epidemiologic relationship between total plasma KB and incident cardiovascular events remains unexplored. In this analysis of a multi-ethnic community-based prospective cohort, we demonstrated that elevated total KB are associated with a range of cardiovascular pathologies and outcomes, including all-cause death, CVD-related death, as well as incident HF. Although the effect sizes were attenuated by adjusting for demographics and a wide range of risk factors, the independent relationship between total KB and cardiovascular outcomes remained statistically significant (Structured Graphical Abstract).
Ketone bodies concentrations can be determined by analyzing urine, blood, or breath. A colorimetric-spectrophotometric assay or NMR-based methods are used to quantify KB in the blood, which are more precise methods.21 Though the evidence on the variability of KB is not entirely known, the limited data suggest that total KB in healthy individuals eating a non-ketogenic diet and engaging in moderate daily physical activity remains relatively stable over time under normal physiologic conditions.24 In healthy adult humans, total KB concentrations can increase in response to physiologic situations such as severe exercise, extended fasting, or adopting a ketogenic diet, as well as pathological diseases such as diabetic ketoacidosis.2 Notably, 95% of the participants had total KB value within a normal range (<0.6 mmol/L), and the extreme conditions mentioned above would increase the level of total KB well above 1 mmol/L. We attempted to account for many of these extraneous factors in our modeling approach. We adjusted for physical activity in our fully adjusted model. We do not believe that any of the participants were affected by starvation, but we did adjust for total calorie intake. In addition, it is unlikely that participants in the MESA study adhered to a ketogenic diet, a nutritional strategy that acquired prominence only recently.
It is currently unknown if individual KB, such as acetoacetate, acetone, and β-OHB, exert a direct role on CVD pathogenesis. A study including 405 stable hemodialysis patients revealed that increased serum β-OHB levels were independently associated with cardiovascular events and all-cause death.25 A population-based Metabolic Syndrome in Men study including 10 106 Finnish men demonstrated that acetoacetate was associated with incident HF.26 Our study demonstrated that both β-OHB and acetoacetate were associated with cardiovascular events and mortality.
Whereas the current clinical evidence regarding the impact of KB on cardiovascular health and disease is inconsistent, our analysis may add clarity to the existing literature. SGLT2i and very low carbohydrate diets increase ketone levels and are becoming increasingly common in clinical practice.27–31 While a metabolic fuel switch to KB utilization has been posited as a potential mechanism of benefit of SGLT2i, these data suggest circulating KB may be associated with cardiovascular events in a broad population. One potential explanation for the seemingly contradictory results may be the effect of time. One could postulate that acute changes in KB might result in a set of specific effects compared with those associated with chronic, lifetime exposure. Furthermore, it is possible that the health state during exposure to elevated KB plays a role in determining the effects of KB. In our analysis, high levels of KB in otherwise healthy subjects were associated with adverse cardiovascular outcomes, while the elevation of KB during a disease state in other settings, such as those with established HF, might offer ‘salvage’ to the failing myocardium by providing an alternative fuel to the diseased tissue.4 Alternatively, the mechanism driving an increase in KB may be relevant to the associated prognostic implication. Therapeutic modulation of KB (for instance, via SGLT2i or very low carbohydrate diets) may be beneficial, while circulating KB (as measured in a general population) may be more reflective of underlying pathological stimulus and thus associated with cardiovascular risk.
Our study is unable to uncover the exact mechanistic explanation underlying KBs apparent association with cardiovascular risk. Prior studies have reported increased myocardial energy expenditure and reduced myocardial efficiency among individuals with elevated KB levels.32 Elevated KB may also capture increased myocardial stress levels that prompt switch in the myocardial substrate to glucose, leading to higher serum ketone bodies levels. It is plausible that ketosis might be a ‘byproduct’ marker of a pre-existing subclinical pathology rather than a direct mediator of cardio-metabolic disorders.25 In our analysis, total KB were associated with incident HF but not with incident stroke or MI. Although the reason for this discrepancy is not entirely apparent, since atherosclerosis is an important underlying mechanism of MI and ischemic stroke, it is plausible to hypothesize that KB exerts its impact via a distinct mechanism on the myocardium. Elevated KB may be physiological and compensatory in patients with heart failure. These hypotheses require further mechanistic examination.
A few studies have shown a positive association between KB and insulin resistance.33–35 In addition, obesity may generate persistent low-grade systemic and local inflammation that contributes to the development of insulin resistance.36 Our analysis demonstrated no evidence of heterogeneity by diabetes or obesity. A prospective population-based cohort study including 6134 participants demonstrated that β-OHB was associated with an increased risk of HF in women but not in men.37 However, our study did not find any significant difference between KB and HF by sex.
Our study found that the correlation of total KB with NT-proBNP and hs-TnT was very poor. NT-proBNP and hs-TnT are established biomarkers of incident HF and myocardial injury.38,39 The orthogonal relations of these biomarkers could be because they are examining different pathophysiologies of disease worsening. In addition, a few studies have demonstrated that insulin resistance is associated with a lower NT-proBNP level.40 Since insulin sensitivity influences ketogenesis, it might be the reason for the lack of correlation. Few studies have demonstrated elevated levels of KB in HF.37,41 However, most of these studies enrolled subjects with CVD at baseline. Exhaled acetone, for instance, was able to identify individuals who had HF with a predictive value that was comparable to that of B-type natriuretic peptide, and the amounts of exhaled acetone positively associated with New York Heart Association class.41 In a recent prospective Dutch cohort of healthy subjects, high levels of β-OHB were associated with an increased risk of developing HF in a sex-specific manner.37 Another study demonstrated that ketone-related metabolites at baseline were predictive of adverse outcomes after coronary artery bypass surgery.42 The findings of our study have important clinical implications. Our study suggests that KB could be used as a potential biomarker, particularly for identifying persons without CVD at baseline who are at the highest risk for incident HF. Further prospective studies to evaluate the role of KB as a biomarker for CVD are desired. The 2021 European Society of Cardiology (ESC) Guidelines on cardiovascular disease prevention also highlight the need for further research in examining the added value of biomarkers in risk classification.43
Several limitations impact the interpretation of our results. The MESA study design necessitates a retrospective analysis of observational data, and thus the causative link between KB and cardiovascular outcomes cannot be determined. Furthermore, repeated measurements of KB were not available; therefore, we were unable to evaluate the association between variability in KB levels and cardiovascular outcomes. Finally, the definition of ‘elevated’ KB in our current study is relative, and overall, the MESA population had KB levels that were generally within what is currently defined as ‘normal range’—therefore, elevated ketones in the MESA cohort do not necessarily reflect the levels observed with ketogenic diets, for instance. However, secondary analyses leveraging restrictive cubic splines showed consistent linear relationships between KB and most of the clinical outcomes. Despite these limitations, the well-established multi-ethnic MESA cohort and its relatively large size strengthen the validity of our results. Given the increasing interest in lifestyle and therapeutic modalities which alter their levels and the lack of a solid body of evidence originating from large-scale analyses to date, additional high-quality data are needed to validate our observations of KB as an emerging cardiovascular biomarker.
Conclusions
Elevated levels of KB in apparently healthy subjects are associated with adverse cardiovascular events, including HF, CVD death, and all-cause death. Our findings demonstrate that KB may serve as a biomarker, particularly for identifying individuals without CVD at baseline who are at the highest risk for incident HF. The addition of plasma KB to a clinical risk score warrants further investigation.
Supplementary Material
Acknowledgements
The authors thank the other investigators, the staff, and the participants of the MESA study for their valuable contributions. A full list of participating MESA investigators and institutions can be found at http://www.mesa-nhlbi.org.
Contributor Information
Elad Shemesh, Institute of Endocrinology, Metabolism and Hypertension, Tel Aviv-Sourasky Medical Center, 6 Weizmann Street, Tel Aviv 6423906, Israel.
Parag Anilkumar Chevli, Section on Hospital Medicine, Department of Internal Medicine, Wake Forest University School of Medicine, Medical Center Blvd, Winston-Salem, NC 27157, USA.
Tareq Islam, Section on Hospital Medicine, Department of Internal Medicine, Geisinger Medical Center, 100 N. Academy Ave, Danville, PA 17822, USA.
Charles A German, Section of Cardiology, Department of Medicine, University of Chicago, 5841 S Maryland Ave, MC 6080, Chicago, IL 60637, USA.
James Otvos, NMR Diagnostics, Morrisville, NC.
Joseph Yeboah, Section on Cardiovascular Medicine, Wake Forest University School of Medicine, Medical Center Blvd, Winston-Salem, NC 27157, USA.
Fatima Rodriguez, Section on Cardiovascular Medicine, Department of Internal Medicine, Stanford University School of Medicine, 300 Pasteur Drive, Stanford, CA 94305, USA.
Christopher deFilippi, Inova Heart and Vascular Institute, 3300 Gallows Rd, Falls Church, VA 22042, USA.
Joao A C Lima, Division of Cardiology, Department of Medicine, Johns Hopkins University, 600 N. Wolfe Street, Baltimore, MD 21287, USA.
Michael Blaha, The Ciccarone Center for the Prevention of Cardiovascular Disease, Johns Hopkins University, 600 N. Wolfe Street, Baltimore, MD 21287, USA.
Ambarish Pandey, Division of Cardiology, Department of Internal Medicine, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390, USA.
Muthiah Vaduganathan, Division of Cardiovascular Medicine, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, MA75 Francis Street, Boston, MA 20115, USA.
Michael D Shapiro, Center for the Prevention of Cardiovascular Disease, Section on Cardiovascular Medicine, Wake Forest University School of Medicine, Medical Center Blvd, Winston-Salem, NC 27157, USA.
Supplementary data
Supplementary data is available at European Heart Journal online.
Pre-registered clinical trial number
None supplied.
Ethical approval statement
Ethical approval was not required.
Data availability statement
Data from the Multi-Ethnic Study of Atherosclerosis (MESA study) can be requested through the National Institutes of Health’s Biologic Specimen and Data Repository Information Coordinating Center (BioLINCC) open program at https://biolincc.nhlbi.nih.gov/studies/mesa/. In addition to the public access repository, interested investigators may also access the data through the MESA Coordinating Center at the University of Washington. Use of the data via this mechanism is overseen by standard MESA policies and procedures, which assure that participant consents are honored and that the topic does not overlap with previously proposed or published work.
Funding statement
The MESA study was funded by contracts 75N92020D00001, HHSN268201500003I, N01-HC-95159, 75N92020D00005, N01-HC-95160, 75N92020D00002, N01-HC-95161, 75N92020D00003, N01-HC-95162, 75N92020D00006, N01-HC-95163, 75N92020D00004, N01-HC-95164, 75N92020D00007, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168, and N01-HC-95169 from the National Heart, Lung, and Blood Institute, and by grants UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 from the National Center for Advancing Translational Sciences (NCATS).
References
- 1. Abdul Kadir A, Clarke K, Evans RD. Cardiac ketone body metabolism. Biochim Biophys Acta Mol Basis Dis 2020;1866:165739. 10.1016/j.bbadis.2020.165739 [DOI] [PubMed] [Google Scholar]
- 2. Puchalska P, Crawford PA. Multi-dimensional roles of ketone bodies in fuel metabolism, signaling, and therapeutics. Cell Metab 2017;25:262–284. 10.1016/j.cmet.2016.12.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Bedi KC Jr, Snyder NW, Brandimarto J, Aziz M, Mesaros C, Worth AJ, et al. Evidence for intramyocardial disruption of lipid metabolism and increased myocardial ketone utilization in advanced human heart failure. Circulation 2016;133:706–716. 10.1161/circulationaha.115.017545 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Aubert G, Martin OJ, Horton JL, Lai L, Vega RB, Leone TC, et al. The failing heart relies on ketone bodies as a fuel. Circulation 2016;133:698–705. 10.1161/circulationaha.115.017355 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Selvaraj S, Kelly DP, Margulies KB. Implications of altered ketone metabolism and therapeutic ketosis in heart failure. Circulation 2020;141:1800–1812. 10.1161/circulationaha.119.045033 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Kolwicz SC, Airhart S, Tian R. Ketones step to the plate A game changer for metabolic remodeling in heart failure? Circulation 2016;133:689–691. 10.1161/Circulationaha.116.021230 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Shimazu T, Hirschey MD, Newman J, He W, Shirakawa K, Le Moan N, et al. Suppression of oxidative stress by beta-hydroxybutyrate, an endogenous histone deacetylase inhibitor. Science 2013;339:211–214. 10.1126/science.1227166 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Gormsen LC, Svart M, Thomsen HH, Sondergaard E, Vendelbo MH, Christensen N, et al. Ketone body infusion with 3-hydroxybutyrate reduces myocardial glucose uptake and increases blood flow in humans: a positron emission tomography study. J Am Heart Assoc 2017;6:e005066. 10.1161/JAHA.116.005066 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Horton JL, Davidson MT, Kurishima C, Vega RB, Powers JC, Matsuura TR, et al. The failing heart utilizes 3-hydroxybutyrate as a metabolic stress defense. JCI Insight 2019;4:e124079. 10.1172/jci.insight.124079 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Nielsen R, Moller N, Gormsen LC, Tolbod LP, Hansson NH, Sorensen J, et al. Cardiovascular effects of treatment with the ketone body 3-hydroxybutyrate in chronic heart failure patients. Circulation 2019;139:2129–2141. 10.1161/Circulationaha.118.036459 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Sattar N, McLaren J, Kristensen S, Preiss D, McMurray JJ. SGLT2 Inhibition and cardiovascular events: why did EMPA-REG outcomes surprise and what were the likely mechanisms? (vol 59, pg 1333, 2016). Diabetologia 2016;59:1573–1574. 10.1007/s00125-016-3987-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Selvaraj S, Fu Z, Jones P, Kwee LC, Windsor SL, Ilkayeva O, et al. Metabolomic profiling of the effects of dapagliflozin in heart failure with reduced ejection fraction: dEFINE-HF. Circulation 2022;146:808–818. 10.1161/circulationaha.122.060402 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Yokokawa T, Sugano Y, Shimouchi A, Shibata A, Jinno N, Nagai T, et al. Exhaled acetone concentration is related to hemodynamic severity in patients with non-ischemic chronic heart failure. Circ J 2016;80:1178–1186. 10.1253/circj.CJ-16-0011 [DOI] [PubMed] [Google Scholar]
- 14. Marcondes-Braga FG, Batista GL, Gutz IGR, Saldiva PHN, Mangini S, Issa VS, et al. Impact of exhaled breath acetone in the prognosis of patients with heart failure with reduced ejection fraction (HFrEF). One year of clinical follow-up. PLoS One 2016;11:e0168790. 10.1371/journal.pone.0168790 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Song JP, Chen L, Chen X, Ren J, Zhang NN, Tirasawasdichai T, et al. Elevated plasma β-hydroxybutyrate predicts adverse outcomes and disease progression in patients with arrhythmogenic cardiomyopathy. Sci Transl Med 2020;12:eaay8329. 10.1126/scitranslmed.aay8329 [DOI] [PubMed] [Google Scholar]
- 16. de Koning MSLY, Westenbrink BD, Assa S, Garcia E, Connelly Margery A, van Veldhuisen Dirk J, et al. Association of circulating ketone bodies with functional outcomes after ST-segment elevation myocardial infarction. J Am Coll Cardiol 2021;78:1421–1432. 10.1016/j.jacc.2021.07.054 [DOI] [PubMed] [Google Scholar]
- 17. Bild DE, Bluemke DA, Burke GL, Detrano R, Roux AVD, Folsom AR, et al. Multi-ethnic study of atherosclerosis: objectives and design. Am J Epidemiol 2002;156:871–881. 10.1093/aje/kwf113 [DOI] [PubMed] [Google Scholar]
- 18. Nettleton JA, Polak JF, Tracy R, Burke GL, Jacobs DR Jr. Dietary patterns and incident cardiovascular disease in the multi-ethnic study of atherosclerosis. Am J Clin Nutr 2009;90:647–654. 10.3945/ajcn.2009.27597 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Levey AS, Stevens LA, Schmid CH, Zhang YP, Castro AF, Feldman HI, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604–612. 10.7326/0003-4819-150-9-200905050-00006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Nguyen K, Fan WJ, Bertoni A, Budoff MJ, Defilippi C, Lombardo D, et al. N-terminal pro B-type natriuretic peptide and high-sensitivity cardiac troponin as markers for heart failure and cardiovascular disease risks according to glucose status (from the multi-ethnic study of atherosclerosis [MESA]). Am J Cardiol 2020;125:1194–1201. 10.1016/j.amjcard.2020.01.025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Garcia E, Shalaurova I, Matyus SP, Oskardmay DN, Otvos JD, Dullaart RPF, et al. Ketone bodies are mildly elevated in subjects with type 2 diabetes Mellitus and are inversely associated with insulin resistance as measured by the lipoprotein insulin resistance Index. J Clin Med 2020;9:321. 10.3390/jcm9020321 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Detrano R, Guerci AD, Carr JJ, Bild DE, Burke G, Folsom AR, et al. Coronary calcium as a predictor of coronary events in four racial or ethnic groups. N Engl J Med 2008;358:1336–1345. 10.1056/NEJMoa072100 [DOI] [PubMed] [Google Scholar]
- 23. Austin PC, Lee DS, Fine JP. Introduction to the analysis of survival data in the presence of competing risks. Circulation 2016;133:601–609. 10.1161/CIRCULATIONAHA.115.017719 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Widjaja A, Morris RJ, Levy JC, Frayn KN, Manley SE, Turner RC. Within- and between-subject variation in commonly measured anthropometric and biochemical variables. Clin Chem 1999;45:561–566. 10.1093/clinchem/45.4.561 [DOI] [PubMed] [Google Scholar]
- 25. Obokata M, Negishi K, Sunaga H, Ishida H, Ito K, Ogawa T, et al. Association between circulating ketone bodies and worse outcomes in hemodialysis patients. J Am Heart Assoc 2017;6:e006885. 10.1161/JAHA.117.006885 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Jauhiainen R, Jauhiainen M, Vangipurapu J, Kuulasmaa T, Ala-Korpela M, Laakso M, et al. Novel biomarkers associated with incident heart failure in 10 106 Finnish men. ESC Heart Fail 2021;8:605–614. 10.1002/ehf2.13132 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. de Cabo R, Mattson MP. Effects of intermittent fasting on health, aging, and disease. N Engl J Med 2019;381:2541–2551. 10.1056/NEJMra1905136 [DOI] [PubMed] [Google Scholar]
- 28. Akinci B. Dapagliflozin and cardiovascular outcomes in type 2 diabetes. N Engl J Med 2019;380:1881. 10.1056/NEJMc1902837 [DOI] [PubMed] [Google Scholar]
- 29. Packer M, Anker SD, Butler J, Filippatos G, Pocock SJ, Carson P, et al. Cardiovascular and renal outcomes with empagliflozin in heart failure. N Engl J Med 2020;383:1413–1424. 10.1056/NEJMoa2022190 [DOI] [PubMed] [Google Scholar]
- 30. Anker SD, Butler J, Filippatos G, Khan MS, Marx N, Lam CSP, et al. Effect of empagliflozin on cardiovascular and renal outcomes in patients with heart failure by baseline diabetes Status: results from the EMPEROR-reduced trial. Circulation 2021;143:337–349. 10.1161/CIRCULATIONAHA.120.051824 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Tragni E, Vigna L, Ruscica M, Macchi C, Casula M, Santelia A, et al. Reduction of cardio-metabolic risk and body weight through a multiphasic very-low calorie ketogenic diet program in women with overweight/obesity: a study in a real-world setting. Nutrients 2021;13:1804. 10.3390/nu13061804 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Du Z, Shen A, Huang Y, Su L, Lai W, Wang P, et al. 1H-NMR-based Metabolic analysis of human serum reveals novel markers of myocardial energy expenditure in heart failure patients. PLoS One 2014;9:e88102. 10.1371/journal.pone.0088102 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Tardif A, Julien N, Pelletier A, Thibault G, Srivastava AK, Chiasson JL, et al. Chronic exposure to beta-hydroxybutyrate impairs insulin action in primary cultures of adult cardiomyocytes. Am J Physiol Endocrinol Metab 2001;281:E1205–E1212. 10.1152/ajpendo.2001.281.6.E1205 [DOI] [PubMed] [Google Scholar]
- 34. Singh BM, Krentz AJ, Nattrass M. Insulin resistance in the regulation of lipolysis and ketone body metabolism in non-insulin dependent diabetes is apparent at very low insulin concentrations. Diabetes Res Clin Pract 1993;20:55–62. 10.1016/0168-8227(93)90023-x [DOI] [PubMed] [Google Scholar]
- 35. Würtz P, Mäkinen VP, Soininen P, Kangas AJ, Tukiainen T, Kettunen J, et al. Metabolic signatures of insulin resistance in 7,098 young adults. Diabetes 2012;61:1372–1380. 10.2337/db11-1355 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Kahn BB, Flier JS. Obesity and insulin resistance. J Clin Invest 2000;106:473–481. 10.1172/jci10842 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Flores-Guerrero JL, Westenbrink BD, Connelly MA, Otvos JD, Groothof D, Shalaurova I, et al. Association of beta-hydroxybutyrate with development of heart failure: sex differences in a Dutch population cohort. Eur J Clin Invest 2021;51:e13468. 10.1111/eci.13468 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Saunders JT, Nambi V, de Lemos JA, Chambless LE, Virani SS, Boerwinkle E, et al. Cardiac troponin T measured by a highly sensitive assay predicts coronary heart disease, heart failure, and mortality in the atherosclerosis risk in communities study. Circulation 2011;123:1367–1376. 10.1161/circulationaha.110.005264 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Di Angelantonio E, Chowdhury R, Sarwar N, Ray KK, Gobin R, Saleheen D, et al. B-Type natriuretic peptides and cardiovascular risk. Circulation 2009;120:2177–2187. 10.1161/CIRCULATIONAHA.109.884866 [DOI] [PubMed] [Google Scholar]
- 40. Khan AM, Cheng S, Magnusson M, Larson MG, Newton-Cheh C, McCabe EL, et al. Cardiac natriuretic peptides, obesity, and insulin resistance: evidence from two community-based studies. J Clin Endocrinol Metab 2011;96:3242–3249. 10.1210/jc.2011-1182 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Marcondes-Braga FG, Gutz IGR, Batista GL, Saldiva PHN, Ayub-Ferreira SM, Issa VS, et al. Exhaled acetone as a new biomaker of heart failure severity. Chest 2012;142:457–466. 10.1378/chest.11-2892 [DOI] [PubMed] [Google Scholar]
- 42. Shah AA, Craig DM, Sebek JK, Haynes C, Stevens RC, Muehlbauer MJ, et al. Metabolic profiles predict adverse events after coronary artery bypass grafting. J Thorac Cardiovasc Surg 2012;143:873–878. 10.1016/j.jtcvs.2011.09.070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Visseren FLJ, Mach F, Smulders YM, Carballo D, Koskinas KC, Bäck M, et al. 2021 ESC guidelines on cardiovascular disease prevention in clinical practice: developed by the task force for cardiovascular disease prevention in clinical practice with representatives of the European Society of Cardiology and 12 medical societies with the special contribution of the European Association of Preventive Cardiology (EAPC). Eur Heart J 2021;42:3227–3337. 10.1093/eurheartj/ehab484 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data from the Multi-Ethnic Study of Atherosclerosis (MESA study) can be requested through the National Institutes of Health’s Biologic Specimen and Data Repository Information Coordinating Center (BioLINCC) open program at https://biolincc.nhlbi.nih.gov/studies/mesa/. In addition to the public access repository, interested investigators may also access the data through the MESA Coordinating Center at the University of Washington. Use of the data via this mechanism is overseen by standard MESA policies and procedures, which assure that participant consents are honored and that the topic does not overlap with previously proposed or published work.