Author + information
- Received January 16, 2018
- Revision received July 22, 2018
- Accepted July 23, 2018
- Published online October 1, 2018.
- Paulo C. Rezende, MD, PhDa,
- Brendan M. Everett, MD, MPHb,
- Maria Mori Brooks, PhDc,
- Helen Vlachos, MSc,
- Trevor J. Orchard, MD, MSc,
- Robert L. Frye, MDd,
- Deepak L. Bhatt, MD, MPHb@DLBhattMD and
- Mark A. Hlatky, MDe,∗ (, )@MarkHlatky_MD@Stanford
- aHeart Institute, University of São Paulo, São Paulo, Brazil
- bDepartment of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts
- cGraduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania
- dDepartment of Medicine, Mayo Clinic, Rochester, Minnesota
- eDepartment of Health Research and Policy, Stanford University School of Medicine, Stanford, California
- ↵∗Address for correspondence:
Dr. Mark A. Hlatky, Stanford University School of Medicine, HRP Redwood Building, Room T150, 259 Campus Drive, Stanford, California 94305-5405.
Background Diabetic medications can cause hypoglycemia, which may lead to myocardial damage.
Objectives This study sought to determine whether hypoglycemia is associated with higher levels of high-sensitivity cardiac troponin T (hsTnT).
Methods The BARI 2D (Bypass Angioplasty Revascularization Investigation 2 Diabetes) trial randomized patients with type 2 diabetes mellitus and stable coronary artery disease, and closely followed them for hypoglycemia over the first year. Hypoglycemia was classified by maximum severity and frequency. hsTnT was measured at baseline and 1 year, and analyzed using multivariable regression.
Results Of 1,984 patients, follow-up hypoglycemia was absent in 1,026 (52%) patients, mild in 875 (44%), and severe in 83 (4%), and occurred less than weekly in 561 (28%) and greater than or equal to weekly in 397 (20%). hsTnT levels were associated with hypoglycemia: a median of 11.4 ng/l (interquartile range [IQR]: 8.1 to 17.3 ng/l) for none, 12.5 ng/l (IQR: 8.3 to 19.3 ng/l) for mild, and 13.7 ng/l (IQR: 9.9 to 24.9 ng/l) for severe hypoglycemia (p = 0.0001); and 12.5 ng/l (IQR: 8.3 to 18.1 ng/l) for less than weekly and 13.0 ng/l (IQR: 8.8 to 21.1 ng/l) for greater than or equal to weekly hypoglycemia (p = 0.0013). Severe hypoglycemia was associated with 34% higher 1-year hsTnT levels (p < 0.0001) in unadjusted analysis, 17% higher (p = 0.006) after adjustment for baseline factors unrelated to diabetes, and 6% higher (p = 0.23) after further adjustment for the duration and severity of diabetes. Hypoglycemia greater than or equal to weekly was associated with 14% higher hsTnT (p = 0.0003) in unadjusted analysis, 12% higher (p = 0.0002) after adjustment for baseline factors unrelated to diabetes, and 4% higher (p = 0.16) after adjustment for diabetes related factors.
Conclusions Hypoglycemia was associated with elevated hsTnT levels, but this may be due to more severe diabetes in patients who developed hypoglycemia, rather than the direct result of hypoglycemia. (Bypass Angioplasty Revascularization Investigation in Type 2 Diabetes [BARI2D]; NCT00006305)
Patients with type 2 diabetes mellitus have an increased risk of developing coronary artery disease, and of dying from heart disease. Stricter control of glucose levels reduces microvascular complications of diabetes, but has had little effect on macrovascular complications, including heart disease. Reducing cardiovascular mortality in patients with diabetes is a public health priority in light of its increasing prevalence worldwide.
Medications that lower glucose levels, particularly insulin and insulin-providing drugs, may produce hypoglycemia. Most episodes of hypoglycemia are mild and readily treated by simple measures, but some episodes can be severe, and may lead to organ dysfunction, especially of the brain. Whether hypoglycemia is also damaging to the heart is less clear. New biomarkers, exemplified by high-sensitivity cardiac-specific troponin assays, can now detect even minor degrees of myocardial damage. Higher levels of circulating troponin have been associated with increased rates of morbidity and mortality (1), even among apparently healthy individuals. The ARIC (Atherosclerosis Risk In Communities) study investigators recently reported an association between troponin levels and a history of hypoglycemia during the previous 20 years among patients with diabetes (2). However, that study was limited by its retrospective design and reliance on administrative data to document hypoglycemia. We sought to determine the association of prospectively documented hypoglycemic episodes with evidence of subclinical myocardial damage among patients with diabetes and coronary artery disease randomized in the BARI 2D (Bypass Angioplasty Revascularization Investigation 2 Diabetes) trial.
The methods and main results of the BARI 2D trial have been described previously (3–5). In brief, between January 1, 2001, and March 31, 2005, 2,368 patients with type 2 diabetes mellitus and stable coronary artery disease were randomized in a 2 × 2 factorial trial to: 1) diabetes management based on either insulin-sensitizing medications or insulin-providing medications; and 2) coronary disease management with either prompt coronary revascularization plus guideline-directed medical therapy or guideline-directed medical therapy alone. The Institutional Review Boards of the University of Pittsburgh and of each participating site approved the trial protocol. All enrolled participants provided written informed consent.
Assessment of hypoglycemia events during first-year follow-up
The clinic staff and study diabetologist prospectively queried patients about any episodes of hypoglycemia at each study visit. Data about the frequency and severity of hypoglycemia were collected during protocol follow-up visits at 1, 2, 3, 4, 5, 6, 9, and 12 months, recorded on a standardized data collection form, and categorized using established criteria. Frequency of hypoglycemia at each time point was graded on a 7-point ordinal scale: none; less than once a month; once a month; more than once a month, but not weekly; once a week; 2 to 4 times a week; or daily. Severity of hypoglycemia was also graded at each assessment based on the characteristics of the episode; specifically, whether it: 1) required assistance of another person; 2) led to calling emergency medical services, visiting the emergency department, or being admitted to the hospital; 3) resulted in a coma or convulsions; or 4) either left the patient unable to self-treat, with a documented blood glucose level <50 mg/dl, or resulted in confusion, irrational behavior, convulsions, or coma reversed by treatment raising blood glucose. Severe hypoglycemia was defined as an episode with any of these severe consequences, whereas mild hypoglycemia was defined as episodes that lacked all of these severe consequences.
Cardiac troponin measurement
A high-sensitivity electrochemiluminescence assay (Roche Diagnostics, Risch-Rotkreuz, Switzerland) was used to measure cardiac troponin T concentration at baseline and at 1 year of follow-up, using plasma samples that had been stored at –80°C. The manufacturer states that this assay has a limit of detection of 3 ng/l, an upper reference limit (99th percentile in healthy volunteers) of 14 ng/l, and a 10% coefficient of variation less than this value. We considered 14 ng/l to be the upper limit of normal, because of its established use in this cohort (1), and in the field. For patients with a high-sensitivity cardiac troponin T (hsTnT) concentration below the lower limit of detection (8 patients at baseline and 4 at 1 year), we imputed a value of 2.9 ng/l.
The study tested the a priori hypothesis that hypoglycemia would be associated with higher hsTnT levels. The primary outcome for this analysis was the concentration of hsTnT at 1-year follow-up. The primary exposure variables for this analysis were the severity and frequency of hypoglycemia episodes during the first year of follow-up, based on the highest category reported at any of the follow-up visits. The 3 severity categories for this analysis were none, mild, and severe, as described previously. The 3 frequency categories for this analysis were none (patients who reported no hypoglycemia episodes during 1 year of follow-up), less than weekly (patients who reported hypoglycemia, with a maximum frequency at any follow-up contact of more than once per month, but not weekly), and greater than or equal to weekly (patients who reported hypoglycemia with a frequency of at least once per week at ≥1 follow-up visits).
We compared the baseline characteristics across the 3 hypoglycemia categories using a Kruskal-Wallis rank-based, nonparametric test for continuous variables and the chi-square test for categorical variables. We also compared hsTnT concentrations at 1 year, and changes from baseline to 1 year, across the 3 hypoglycemia categories using a Kruskal-Wallis test.
We used a series of linear regression models to assess the association of hypoglycemia with 1-year hsTnT levels. As the troponin levels were skewed, we used a natural logarithmic transformation of 1-year hsTnT values as the dependent variable in all regression models. In the first regression analysis, we used only the hypoglycemia categories as independent variables with no hypoglycemia as the reference category. In the second step, we further adjusted for baseline hsTnT, using its natural logarithmic transformed value. In the next model, we additionally adjusted for demographic factors (age, race, sex), established cardiac risk factors (systolic blood pressure, total and high-density lipoprotein cholesterol, body mass index, and estimated glomerular filtration rate), and severity of coronary heart disease (history of myocardial infarction, history of heart failure, number of diseased coronary arteries, and left ventricular ejection fraction <50%). In the final model, we further adjusted for diabetes-related factors at baseline (duration of diabetes, glycosylated hemoglobin [HbA1c] levels, use of insulin, and history at baseline of prior hypoglycemia).
We performed sensitivity analyses using stepwise regression to determine which baseline diabetes-related factors had the greatest effect on the association between hypoglycemia and 1-year hsTnT levels, controlling for baseline hsTnT levels.
Finally, we performed an intention-to-treat analysis that compared 1-year hsTnT levels between patients randomly assigned to the insulin provision strategy or the insulin sensitization strategy, using the Wilcoxon rank sum test. All analyses were performed using SAS version 9.3 (SAS Institute, Cary, North Carolina).
Of the 2,368 patients randomized in the BARI 2D trial, 2,285 (96.5%) had hsTnT concentrations measured at baseline. After 1 year of follow-up, 66 patients (2.9%) had died, primarily due to cardiovascular disease (n = 43) or diabetes complications (n = 17); 1 patient died of severe hypoglycemia (Online Table 1). A further 235 (10.3%) patients did not have 1-year biomarker measurements available. The 1,984 patients who had both baseline and 1-year follow-up hsTnT measurements, as well as follow-up data on hypoglycemia, formed the analytic cohort for the present study (Figure 1).
During the first year of trial treatment and follow-up, 1,026 (52%) patients had no episodes of hypoglycemia, 875 (44%) patients had only mild hypoglycemia, and 83 (4%) patients had 1 or more episodes of severe hypoglycemia. Patients with no hypoglycemia were significantly more likely to be men, have higher diastolic blood pressure, have had a shorter duration of diabetes, and have lower levels of glycated hemoglobin and insulin use at baseline (Table 1). Patients with more severe hypoglycemia were more likely to have been randomly assigned to the insulin provision strategy, and to have previously used insulin (Table 1), and to be taking insulin during follow-up (Online Table 2).
The frequency of hypoglycemia episodes varied among the 958 patients who had at least 1 episode during the first year: 561 patients had hypoglycemia episodes less than weekly; 157 less than monthly; 156 once monthly; 248 several times a month, but less than weekly; and 397 greater than or equal to weekly (142 weekly, 233 several times a week, 22 daily). Patients with more frequent hypoglycemia were more likely to be women, have lower diastolic blood pressure levels, have a longer duration of diabetes, have higher glycated hemoglobin levels, have greater use of insulin at baseline (Table 2), and be taking insulin during follow-up (Online Table 3). Patients with more frequent hypoglycemia episodes were more likely to have been randomly assigned to the insulin-providing strategy and have had higher concentrations of hsTnT at baseline (Table 2).
1-year troponin T analyses
Levels of hsTnT at 1-year follow-up were significantly higher (p < 0.0001) among patients who had hypoglycemia events during the first year of follow-up (Figure 2): the median 1-year hsTnT levels were 11.4 ng/l (interquartile range [IQR]: 8.1 to 17.3 ng/l) for patients without follow-up hypoglycemia, 12.5 ng/l (IQR: 8.3 to 19.3 ng/l) for patients with mild hypoglycemia, and 13.7 ng/l (IQR: 9.9 to 24.9 ng/l) for patients with severe hypoglycemia (p = 0.0001). One-year hsTnT levels were above normal (>14 ng/l) significantly more often in patients with hypoglycemia (p = 0.0018): 363 (35%) patients without hypoglycemia, 367 (42%) patients with mild hypoglycemia, and 41 (49%) patients with severe hypoglycemia had elevated hsTnT levels. The changes in hsTnT levels from baseline to 1 year were also significantly larger (p = 0.038) among patients with hypoglycemia, with median changes of 0.6 ng/l (IQR: –1.3 to +2.7 ng/l) for no hypoglycemia, 0.9 ng/l (IQR: –1.2 to +3.3 ng/l) for mild hypoglycemia, and 2.0 ng/l (IQR: –1.2 to +5.2 ng/l) for severe hypoglycemia.
More frequent episodes of hypoglycemia during the first year of follow-up were also significantly associated (p = 0.001) with higher levels of hsTnT at 1 year (Figure 3): the median 1-year hsTnT levels were 11.4 ng/l (IQR: 8.1 to 17.3 ng/l) for patients with no hypoglycemia, 12.5 ng/l (IQR: 8.3 to 18.1 ng/l) for those with hypoglycemia less than weekly, and 13.0 ng/l (IQR: 8.8 to 21.1 ng/l) for those with hypoglycemia greater than or equal to weekly. One-year hsTnT levels were above the upper limit of normal in 363 (35%) patients without hypoglycemia, 230 (41%) patients with episodes less than weekly, and 178 (45%) patients with episodes greater than or equal to weekly. The changes in hsTnT levels from baseline to 1 year were larger, but not significantly so, among those with more frequent hypoglycemia: median of 0.6 ng/l (IQR: –1.3 to +2.7 ng/l) in those without hypoglycemia, 1.0 ng/l (IQR: –1.3 to +3.3 ng/l) in those with episodes less than weekly, and 1.0 ng/l (IQR: –1.0 to +3.4 ng/l) among those with episodes greater than or equal to weekly (p = 0.075).
We used regression models to assess the association of 1-year hsTnT levels with the severity and frequency of hypoglycemia, first unadjusted, then adjusted progressively for various baseline factors. In an unadjusted model, severity of hypoglycemia was strongly associated (p = 0.0001) with 1-year hsTnT levels, as was frequency of hypoglycemia (p = 0.001) (Table 3). These associations were attenuated, but remained significant, after adjustment for baseline hsTnT levels. The associations were essentially unchanged by further adjustment for baseline factors that were unrelated to diabetes (Table 3). However, the associations were greatly attenuated, and no longer significant, after adjustment for baseline factors related to diabetes: diabetes duration, baseline HbA1c, baseline insulin use, and a baseline history of hypoglycemia (Table 3). When diabetes related factors were added stepwise to models that adjusted only for baseline hsTnT levels, the associations of hypoglycemia with 1-year hsTnT levels were no longer significant after adjustment for the duration of diabetes (Table 4, Online Figure 1).
Patients randomly assigned to the insulin provision strategy were more likely to have episodes of hypoglycemia during follow-up (58% vs. 39%; p < 0.0001). Compared with those assigned to insulin sensitization, mild hypoglycemia was more common (53% vs. 35%), as was severe hypoglycemia (4.9% vs. 3.5%; p < 0.0001). Episodes of less than weekly hypoglycemia were more common in patients assigned to the insulin provision strategy (32% vs. 25%), as were episodes of greater than or equal to weekly hypoglycemia (26% vs. 14%; p < 0.0001).
However, 1-year hsTnT levels did not differ significantly between patients assigned to insulin provision compared with insulin sensitization (p = 0.68). The median hsTnT was 12.1 ng/l (IQR: 8.4 to 17.5 ng/l) among patients assigned to the insulin provision strategy, and 11.9 ng/l (IQR: 8.0 to 18.9 ng/l) among patients randomly assigned to the insulin sensitization strategy. The change in hsTnT levels from baseline to 1 year also did not differ significantly according to random assignment to insulin provision or insulin sensitization: median change 0.7 ng/l (IQR: –1.2 to +2.8 ng/l) versus 0.9 ng/l (IQR: –1.3 to +3.4 ng/l) (p = 0.25).
Hypoglycemia episodes documented prospectively over 1 year of follow-up among patients with diabetes and coronary artery disease were significantly associated with higher levels of hsTnT, a marker of subclinical myocardial damage and of increased cardiovascular risk (Central Illustration). This association remained significant after adjustment for multiple baseline factors, including demographics, severity of coronary disease, and initial levels of hsTnT. The association was attenuated and no longer significant, however, after adjustment for baseline measures of diabetes duration and severity, which suggests that hypoglycemia per se may not have caused the elevated hsTnT levels, but instead may identify patients with more severe diabetes. Our data, although suggestive, cannot determine whether episodes of hypoglycemia lead directly to higher levels of hsTnT, or are merely associated with them.
The pathophysiologic processes that lead to elevated levels of hsTnT among apparently stable, healthy patients have not been established. This study shows that diabetes-related factors, including hypoglycemia, are significantly associated with higher hsTnT levels in a graded fashion. These various measures of diabetes are correlated with one another, which makes it difficult to disentangle their potentially independent effects on hsTnT, and to distinguish mediation from confounding. An episode of hypoglycemia might lead to an immediate rise in hsTnT levels, but our study was not designed to detect this. Patients with recorded episodes of hypoglycemia may well have had additional, recurrent episodes that could contribute to chronic increases in hsTnT. The duration of diabetes, which we found to be most strongly associated with hsTnT levels, is associated with end-organ damage, and with greater use of diabetes medications, particularly insulin, which may lead to hypoglycemia. As the measures of diabetes severity were strongly associated with each other and with hypoglycemia, it remains possible that hypoglycemia might be part of a causal pathway from diabetes to subclinical myocardial damage and elevated hsTnT levels.
Hypoglycemia activates the sympathoadrenal system, and leads to the release of catecholamines. These effects may be especially deleterious to patients with established coronary disease when myocardial energy supply is reduced due to hypoglycemia. Several clinical research studies suggest that hypoglycemia may indeed lead to myocardial ischemia. Ambulatory electrocardiogram recordings among patients with type 2 diabetes and documented coronary disease showed that episodes of hypoglycemia were associated with acute ischemic electrocardiogram changes and angina (6). Hypoglycemia has also been shown to affect cardiac repolarization, leading to altered ST-segment elevation and T-wave morphology and prolongation of QT interval (7,8), and may have procoagulant and proinflammatory effects (9). These consequences of hypoglycemia may be deleterious to the cardiovascular system, and lead to higher troponin levels.
Although mechanistic studies suggest that hypoglycemia might lead to cardiovascular events, the evidence from large prospective studies is limited, in part because few deaths or myocardial infarctions are directly due to hypoglycemia. In the ACCORD (Action to Control Cardiovascular Risk in Diabetes) trial (10), patients with diabetes and established cardiovascular disease or elevated risk were randomized to an intensive (HbA1c <6.0%) or standard (HbA1c 7.0% to 7.9%) glucose targets. The ACCORD trial was halted prematurely because of higher cardiac mortality in the intensive arm group. Subsequent analyses (11) indicated that patients who experienced hypoglycemia had higher adjusted hazard ratios for mortality, but few deaths were caused by severe hypoglycemia. Similarly, analyses from the ORIGIN (Outcomes Reduction with an Initial Glargine Intervention) (12) and ADVANCE (Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation) studies (13) have also found associations between hypoglycemia episodes and higher risk of major cardiovascular events. However, these studies could not determine whether hypoglycemia was the cause of cardiovascular complications, especially because there was no clear temporal relationship between hypoglycemic episodes and clinical events.
The only previous study that has specifically addressed the association between hypoglycemia and cardiac troponin levels was performed by the ARIC study investigators (2). They reported that 3% of 2,148 participants had a documented episode of severe hypoglycemia in the prior 20 years, and that these patients had higher troponin T levels. The association remained significant after they adjusted for covariates such as age, race, gender, and history of coronary heart disease and heart failure, but was no longer significant after adjustments for HbA1c levels, diabetes duration, and estimated glomerular filtration rate. The ARIC study was limited by its ascertainment of hypoglycemia episodes by administrative diagnosis codes, and the long intervals between the hypoglycemic event and the measurement of troponin levels (median of 4.3 years). The overall findings of the ARIC study are similar to ours: we both found a strong association of hypoglycemia with elevated troponin levels that was attenuated by adjustment for diabetes-related factors.
Studies from general populations suggest that diabetes is independently associated with higher troponin levels, and that patients with diabetes with higher troponin levels are more likely to develop cardiovascular events during follow-up (1,14,15). Another study of patients with diabetes (16) confirmed that higher levels of troponin were associated with higher rates of cardiovascular events, and reported that glomerular filtration rate and the presence of coronary disease were independent predictors of troponin levels, but diabetes-related factors were not, including diabetes duration, levels of HbA1c and glucose, and use of insulin.
The evidence suggests that troponin levels are predictive of future adverse cardiac events—in the general population, in patients with diabetes, and in patients with coronary artery disease. The findings of our study add to this evidence by showing that episodes of hypoglycemia are associated with higher troponin levels, although this association appears to be closely tied to other markers of the severity of diabetes.
Although we assessed hypoglycemia prospectively at multiple time points during a year of follow-up, hsTnT was measured only twice—at baseline and at 1-year follow-up. Therefore, we could not assess the trajectory of changes in hsTnT over follow-up, or whether specific episodes of hypoglycemia led to immediate increases in hsTnT levels. The BARI 2D trial did not collect data on self-monitored glucose levels, and we were unable to assess those levels systematically. The trial recorded data on symptoms of confusion, coma, or convulsions due to hypoglycemia, but did not record data on other symptoms. Patients in the BARI 2D trial were selected, and closely managed, and therefore may not be representative of the patients with diabetes seen in general practice. Finally, we do not have data on changes in hsTnT among the 66 patients who died before the end of 1-year follow-up (Online Table 1).
We found that episodes of hypoglycemia were associated with significantly higher levels of hsTnT among patients with diabetes and coronary disease. This finding suggests that hypoglycemia may be a marker of ongoing subclinical myocardial damage in patients with diabetes, but the mechanism(s) underlying this association is uncertain, and may be due to a constellation of factors related to the severity of diabetes.
COMPETENCY IN MEDICAL KNOWLEDGE: In patients with type 2 diabetes mellitus and stable coronary artery disease, hypoglycemia is associated with elevated levels of troponin-T measured by high-sensitivity assay, reflecting either more severe diabetes or a direct adverse effect of hypoglycemia.
TRANSLATIONAL OUTLOOK: Future research should explore the mechanisms linking hypoglycemia and elevated cardiac troponin levels and assess the prognostic implications of this association.
The authors thank Thomas Sehested for producing the figures.
Supported by grants from the National Heart, Lung, and Blood Institute, the National Institute of Diabetes and Digestive and Kidney Diseases and Roche Diagnostics (U01HL061744, U01HL061746, U01HL061748, U01HL063804, and R21HL121495). The study sponsors had no role in the design, data collection, data analysis, or interpretation of this study, and no role in writing this report or the decision to submit the manuscript. Dr. Everett has received grant support from Novartis Pharmaceuticals; and has served as a consultant for Roche Diagnostics and Abbott Laboratories. Dr. Bhatt has served on the advisory board for Cardax, Elsevier Practice Update Cardiology, Medscape Cardiology, and Regado Biosciences; has served on the Board of Directors of the Boston VA Research Institute and Society of Cardiovascular Patient Care; has served as the chair of American Heart Association Quality Oversight Committee; has served on the data monitoring committees of the Cleveland Clinic, Duke Clinical Research Institute, Harvard Clinical Research Institute, Mayo Clinic, and Population Health Research Institute; has received honoraria from the American College of Cardiology (Senior Associate Editor, Clinical Trials and News, ACC.org), Belvoir Publications (Editor in Chief, Harvard Heart Letter), Duke Clinical Research Institute (clinical trial steering committees), Harvard Clinical Research Institute (clinical trial steering committee), HMP Communications (Editor in Chief, Journal of Invasive Cardiology), Journal of the American College of Cardiology (Guest Editor; Associate Editor), Population Health Research Institute (clinical trial steering committee), Slack Publications (Chief Medical Editor, Cardiology Today Intervention), Society of Cardiovascular Patient Care (Secretary/Treasurer), and WebMD (CME steering committees); has served as the Deputy Editor for Clinical Cardiology; has served as the chair of the NCDR-ACTION Registry Steering Committee and VA CART Research and Publications Committee; has received research funding from Amarin, Amgen, AstraZeneca, Bristol-Myers Squibb, Chiesi, Eisai, Ethicon, Forest Laboratories, Ironwood, Ischemix, Lilly, Medtronic, Pfizer, Roche, Sanofi, and The Medicines Company; has received royalties from Elsevier (Editor, Cardiovascular Intervention: A Companion to Braunwald Heart Disease); has served as a site co-investigator for Biotronik, Boston Scientific, and St. Jude Medical (now Abbott); has served as a trustee of American College of Cardiology; and has performed unfunded research for FlowCo, PLx Pharma, and Takeda. Dr. Hlatky has served as the Medical Advisor for the Office of Clinical Affairs of the Blue Cross Blue Shield Association; served an associate editor for Journal of the American College of Cardiology; served a consultant to the George Institute and Acumen, Inc.; and received research funding from HeartFlow, Inc., Milestone Pharmaceuticals, St. Jude Medical, and Sanofi. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- glycosylated hemoglobin
- high-sensitivity cardiac troponin T
- interquartile range
- Received January 16, 2018.
- Revision received July 22, 2018.
- Accepted July 23, 2018.
- 2018 American College of Cardiology Foundation
- Lee A.K.,
- McEvoy J.W.,
- Hoogeveen R.C.,
- Ballantyne C.M.,
- Selvin E.
- Desouza C.,
- Salazar H.,
- Cheong B.,
- Murgo J.,
- Fonseca V.
- Tsujimoto T.,
- Yamamoto-Honda R.,
- Kajio H.,
- et al.
- Wright R.J.,
- Newby D.E.,
- Stirling D.,
- Ludlam C.A.,
- Macdonald I.A.,
- Frier B.M.
- Bonds D.E.,
- Miller M.E.,
- Bergenstal R.M.,
- et al.
- Wallace T.W.,
- Abdullah S.M.,
- Drazner M.H.,
- et al.
- Everett B.M.,
- Cook N.R.,
- Magnone M.C.,
- et al.