OBJECTIVE

The glucagon-like peptide-1 receptor agonist dulaglutide reduced MACE in the Researching Cardiovascular Events with a Weekly Incretin in Diabetes (REWIND) trial. This article expores the relationship of selected biomarkers to both dulaglutide and major adverse cardiovascular events (MACE).

RESEARCH DESIGN AND METHODS

In this post hoc analysis, stored fasting baseline and 2-year plasma samples from 824 REWIND participants with MACE during follow-up and 845 matched non-MACE participants were analyzed for 2-year changes in 19 protein biomarkers. Two-year changes in 135 metabolites were also analyzed in 600 participants with MACE during follow-up and in 601 matched non-MACE participants. Linear and logistic regression models were used to identify proteins that were associated with both dulaglutide treatment and MACE. Similar models were used to identify metabolites that were associated with both dulaglutide treatment and MACE.

RESULTS

Compared with placebo, dulaglutide was associated with a greater reduction or lesser 2-year rise from baseline in N-terminal prohormone of brain natriuretic peptide (NT-proBNP), growth differentiation factor 15 (GDF-15), high-sensitivity C-reactive protein, and a greater 2-year rise in C-peptide. Compared with placebo, dulaglutide was also associated with a greater fall from baseline in 2-hydroxybutyric acid and a greater rise in threonine (P < 0.001). Increases from baseline in two of the proteins (but neither metabolite) were associated with MACE, including NT-proBNP (OR 1.267; 95% CI 1.119, 1.435; P < 0.001) and GDF-15 (OR 1.937; 95% CI 1.424, 2.634; P < 0.001).

CONCLUSIONS

Dulaglutide was associated with a reduced 2-year rise from baseline of NT-proBNP and GDF-15. Higher rises of these biomarkers were also associated with MACE.

People with type 2 diabetes are at high risk for cardiovascular (CV) outcomes and premature death. Although the exact reasons for this increased risk remain elusive, large CV outcomes trials have identified a variety of therapeutic approaches to reduce the incidence of these outcomes. The glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are one such class (1). In addition to lowering glucose, weight, and blood pressure in people with type 2 diabetes, meta-analyses of nine large randomized controlled trials (2,3) have reported that GLP-1 RAs reduce the risk of major adverse CV events (MACE), defined as a nonfatal myocardial infarction, nonfatal stroke, or death from CV or unknown causes. Although concomitant changes in glycemic control and albuminuria may account for some of these CV benefits (4,5), much less is known regarding the relationship between treatment with these drugs and treatment-related changes in the circulating protein and metabolite biomarkers that are also associated with MACE.

The Researching Cardiovascular Events with a Weekly Incretin in Diabetes (REWIND) trial reported that the GLP-1 RA dulaglutide reduced MACE in middle-aged people with type 2 diabetes and either previous CV disease (CVD) or CV risk factors (6). The availability of stored biological samples in a subset of REWIND participants at baseline and postrandomization provided an opportunity to identify protein and metabolite biomarkers that are associated with both dulaglutide assignment and MACE outcomes (7,8).

The design, main results, and additional analyses of the international REWIND trial were previously published (6,9,10). Briefly, 9,901 middle-aged and older people (mean age 66 years) with type 2 diabetes (46% women) and additional CV risk factors were recruited. Participants were randomly assigned to the addition of weekly subcutaneous injections of either 1.5 mg dulaglutide or placebo and followed for a median of 5.4 years for the occurrence of MACE and other health outcomes. A subset of participants provided blood samples at both the baseline and 2-year visit for storage and subsequent biomarker analyses. The trial was approved by ethics committees at all participating sites, and all participants provided written informed consent that included consent to store and analyze blood samples.

Sample Preparation and Storage and Analysis for Proteins and Metabolites

Fasting serum and EDTA plasma specimens were collected and spun, separated, and aliquoted into two serum and two plasma aliquots at the recruitment sites, where they were frozen at −20°C and shipped to the central lab(ICON) on dry ice for storage at −80°C. Proteins were measured in duplicate using a variety of assays (Supplementary Tables 1 and 2), normalized to standard curves, and expressed as the averaged concentrations of the two measurements. Metabolites were assayed using a targeted liquid chromatography–tandem mass spectrometry omics approach, normalized to a pooled sample standard curve, and expressed as relative percentages (Supplementary Materials and Methods).

Choice of Cases and Controls

Biomarkers were analyzed using a nested case-control design, with cases being any individual with stored plasma samples from the baseline and 2-year visits who experienced a MACE during follow-up. Controls were defined as individuals with stored baseline and 2-year plasma who did not experience a MACE postrandomization. One control was matched to each case based on treatment assignment, sex, and baseline statin use, blood pressure, LDL cholesterol, and HbA1c within 1 SD of the values for all biomarker study participants; age within 5 years; and a follow-up time in the trial that was longer than the follow-up time of the corresponding case. Controls for cases that remained unmatched using the above criteria were identified based on a blood pressure, LDL cholesterol, and HbA1c within 2 SD, and age within 10 years.

Statistical Analysis

All statistical analyses were performed using SAS version 9.4. Categorical variables were summarized using counts with percentages and compared using χ2 tests. Continuous clinical variables were summarized using means with SD and compared using t tests. Protein concentrations and metabolite percentages were analyzed as continuous variables after natural logarithm (ln) transformation.

Protein analyses were restricted to the 19 proteins for which baseline and 2-year levels were above the lower limit of detection in >95% of the case-control samples. Values below the lower limit of detection were imputed using a truncated normal distribution in which the mean was half of the lower limit of detection and in which variability was similar to that of the detectable levels. Metabolite analyses were restricted to 135 metabolites detected in >75% of individual samples. Metabolite percentages were estimated from a pooled reference sample. For each protein and metabolite biomarker, the variable used for all analyses was the difference between the ln (2-year) and ln (baseline) levels, calculated as ln (2-year/baseline). The analyses were conducted in two stages (Fig. 1). First, they identified biomarkers that were independently associated with dulaglutide. Then they determined which of the identified biomarkers were independently associated with MACE.

A discovery approach was first used to identify biomarkers that were independently associated with random assignment to dulaglutide versus placebo. This was done using one linear regression model for each biomarker measured, in which the dependent variable was ln (2-year/baseline) and the independent variables were dulaglutide assignment as well as ln (baseline biomarker level), age, sex, White ancestry, systolic blood pressure, previous CV disease, current smoking, BMI, HbA1c, the estimated glomerular filtration rate (eGFR), LDL cholesterol, and albuminuria. To control type 1 errors arising from multiple statistical tests, only those models for which the P value for the dulaglutide term was less than the Bonferroni-adjusted P value (i.e., 0.05 divided by the number of estimated models) were deemed to identify a statistically significant dulaglutide-linked biomarker change.

Biomarker changes linked with dulaglutide that were identified using the discovery approach were then assessed to determine whether they were also independently associated with MACE. This was done using one logistic regression model for each biomarker, in which the dependent variable was MACE, and the independent variables were ln (2-year/baseline) for each biomarker as well as age, sex, White self-identification, systolic blood pressure, previous CV disease, current smoking, BMI, HbA1c, eGFR, LDL cholesterol, and albuminuria. In separate models, dulaglutide assignment as well as an interaction term of dulaglutide × ln (2-year/baseline) were included. Only those models for which the P value for the biomarker change term was <0.05 divided by the number of estimated models were deemed to identify a statistically significant MACE-linked biomarker change.

Biomarkers were assayed in stored samples from 1,809 participants, 902 who experienced a MACE postrandomization and 907 non-MACE controls who were matched to cases using the two-step approach described in Research Design and Methods. Of the 1,809 participants, a subset of 1,669 (824 MACE and 845 non-MACE) had recorded levels for all 19 protein biomarkers (Supplementary Table 1) at both the baseline and 2-year visit in >95% of participants (with imputed values for the remainder). In addition, a subset of 1,201 of the 1,809 participants (600 MACE, 601 non-MACE) had recorded measurements for 135 metabolite biomarkers (Supplementary Table 3) at baseline and 2 years. In both the protein biomarker and metabolite biomarker analyses, the mean participant age was 67 years, 37% were female, 15% were smokers, and 38% had prior CVD. These and other baseline characteristics of participants included in the protein and metabolite analyses are noted in Table 1. This table also shows a balanced distribution of baseline characteristics in participants assigned to either dulaglutide or placebo for both the protein and metabolite analyses.

Table 1

Characteristics of participants in the biomarker studies

Protein biomarkersMetabolite biomarkers
DulaglutidePlaceboPDulaglutidePlaceboP
N 797 872  596 605  
Age (years) 67.6 (6.7) 67.2 (6.9) 0.31 67.4 (6.7) 66.8 (7.0) 0.12 
Females (%) 287 (36.0%) 332 (38.1) 0.38 220 (36.9) 224 (37.0) 0.97 
White (%) 618 (77.5) 664 (76.2) 0.50 466 (78.2) 456 (75.4) 0.25 
Current smoking (%) 128 (16.1) 122 (14.0) 0.21 91 (15.3) 88 (14.6) 0.69 
Previous CVD (%) 300 (37.6) 330 (37.8) 0.95 221 (37.1) 236 (39.0) 0.54 
BMI (kg/m232.1 (5.6) 32.0 (5.5) 0.64 32.2 (5.7) 32.3 (5.5) 0.79 
Systolic BP (mm Hg) 139 (15.9) 140 (17.6) 0.47 139 (16.0) 139 (17.6) 0.79 
Diastolic BP (mm Hg) 78.6 (9.89) 78.9 (10.3) 0.61 78.3 (9.9) 78.6 (10.0) 0.63 
eGFR (mL/min/1.73 m275.3 (22.9) 76.2 (22.6) 0.38 76.4 (23.0) 76.9 (22.4) 0.73 
Albuminuria (%) 314 (39.4) 358 (41.1) 0.41 239 (40.1) 247 (40.8) 0.69 
HbA1c (%) 7.0 (1.0) 7.0 (1.0) 0.71 7.4 (1.0) 7.3 (1.0) 0.81 
LDL cholesterol (mmol/L) 2.55 (0.95) 2.58 (0.97) 0.46 2.53 (0.96) 2.56 (0.99) 0.53 
MACE cases (%) 392 (49.2) 432 (49.5) 0.88 298 (50.0) 302 (49.9) 0.98 
Controls (%) 405 (50.8) 440 (50.5) 0.88 298 (50.0) 303 (50.1) 0.98 
Protein biomarkersMetabolite biomarkers
DulaglutidePlaceboPDulaglutidePlaceboP
N 797 872  596 605  
Age (years) 67.6 (6.7) 67.2 (6.9) 0.31 67.4 (6.7) 66.8 (7.0) 0.12 
Females (%) 287 (36.0%) 332 (38.1) 0.38 220 (36.9) 224 (37.0) 0.97 
White (%) 618 (77.5) 664 (76.2) 0.50 466 (78.2) 456 (75.4) 0.25 
Current smoking (%) 128 (16.1) 122 (14.0) 0.21 91 (15.3) 88 (14.6) 0.69 
Previous CVD (%) 300 (37.6) 330 (37.8) 0.95 221 (37.1) 236 (39.0) 0.54 
BMI (kg/m232.1 (5.6) 32.0 (5.5) 0.64 32.2 (5.7) 32.3 (5.5) 0.79 
Systolic BP (mm Hg) 139 (15.9) 140 (17.6) 0.47 139 (16.0) 139 (17.6) 0.79 
Diastolic BP (mm Hg) 78.6 (9.89) 78.9 (10.3) 0.61 78.3 (9.9) 78.6 (10.0) 0.63 
eGFR (mL/min/1.73 m275.3 (22.9) 76.2 (22.6) 0.38 76.4 (23.0) 76.9 (22.4) 0.73 
Albuminuria (%) 314 (39.4) 358 (41.1) 0.41 239 (40.1) 247 (40.8) 0.69 
HbA1c (%) 7.0 (1.0) 7.0 (1.0) 0.71 7.4 (1.0) 7.3 (1.0) 0.81 
LDL cholesterol (mmol/L) 2.55 (0.95) 2.58 (0.97) 0.46 2.53 (0.96) 2.56 (0.99) 0.53 
MACE cases (%) 392 (49.2) 432 (49.5) 0.88 298 (50.0) 302 (49.9) 0.98 
Controls (%) 405 (50.8) 440 (50.5) 0.88 298 (50.0) 303 (50.1) 0.98 

Data are means (SD) shown for continuous variables. The characteristics of participants with recorded levels for all biomarkers at the baseline and 2-year visit are shown. Previous CVD denotes previous myocardial infarction, ischemic stroke, unstable angina with electrocardiogram changes, myocardial ischemia on imaging or stress test, or coronary, carotid, or peripheral revascularization. MACE denotes a nonfatal myocardial infarction, nonfatal stroke, or death from CV or unknown causes. Albuminuria, urine albumin/creatinine ≥ 3.39 mg/mmol; BP, blood pressure.

Relationship Between the 2-Year Change in Protein Biomarkers, Dulaglutide Assignment, and MACE

Nineteen separate adjusted linear regression models analyzed the relationship between ln (2-year/baseline) for each protein biomarker that was measured and the random assignment of participants to dulaglutide versus placebo. These identified four protein biomarkers (Supplementary Table 4) for which the change from baseline was significantly associated with dulaglutide based on a Bonferroni-corrected P value of 0.0026 (i.e., 0.05/19). These biomarkers were high-sensitivity C-reactive protein (hsCRP), N-terminal pro brain natriuretic peptide (NTproBNP), growth differentiation factor 15 (GDF15), and C-peptide. As noted in Fig. 2A, dulaglutide assignment was associated with a lesser increase or greater reduction from baseline than placebo for hsCRP, NT-proBNP, and GDF15, and a greater increase from baseline than placebo for C-peptide.

Figure 1

Analytic approach used to conduct the biomarker analyses. No., number.

Figure 1

Analytic approach used to conduct the biomarker analyses. No., number.

Close modal
Figure 2

Adjusted change in protein (A) or metabolite (B) level in people randomly assigned to dulaglutide or placebo. Positive values denote a rise from baseline to 2 years.

Figure 2

Adjusted change in protein (A) or metabolite (B) level in people randomly assigned to dulaglutide or placebo. Positive values denote a rise from baseline to 2 years.

Close modal

The odds of MACE per unit increase from baseline (i.e., per unit increase in ln [2 year/baseline]) for each of the four proteins identified above was estimated with adjusted logistic regression models. As noted in Table 2, the CIs for the odds ratios excluded one for hsCRP (OR 1.138: 95% CI 1.024, 1.265; P = 0.016), NT-proBNP (OR 1.267: 95% CI 1.119, 1.435; P < 0.001), and GDF15 (OR 1.937, 95% CI 1.424, 2.634; P < 0.001). Two of these biomarkers (NT-proBNP and GDF15) were significantly associated with case status based on a Bonferroni-corrected P value < 0.0125 (i.e., 0.05/4 comparisons). None of the biomarkers interacted with dulaglutide assignment with respect to MACE. In a sensitivity analysis that excluded the 385 cases whose MACE occurred after the 2-year visit (Supplementary Table 5), a significant association between MACE and changes in GDF15 and hsCRP but not NT-proBNP was noted.

Table 2

Association between MACE status and change in specific biomarker levels from baseline

Adjusted for dulaglutide and risk factors OR per unit higher level (95% CI)P
Proteins   
 High-sensitivity CRP 1.138 (1.024, 1.265) 0.016 
 NT-proBNP 1.267 (1.119, 1.435) <0.001 
 GDF15 1.937 (1.424, 2.634) <0.001 
 C-peptide 1.130 (0.981, 1.302) 0.090 
Metabolites   
 2-HB 1.150 (0.840, 1.573) 0.38 
 Threonine 1.063 (0.566, 1.994) 0.85 
Adjusted for dulaglutide and risk factors OR per unit higher level (95% CI)P
Proteins   
 High-sensitivity CRP 1.138 (1.024, 1.265) 0.016 
 NT-proBNP 1.267 (1.119, 1.435) <0.001 
 GDF15 1.937 (1.424, 2.634) <0.001 
 C-peptide 1.130 (0.981, 1.302) 0.090 
Metabolites   
 2-HB 1.150 (0.840, 1.573) 0.38 
 Threonine 1.063 (0.566, 1.994) 0.85 

The relationship between a one-unit higher ln (2-year) – ln (baseline) and MACE status is shown. Risk factors included in the model are age, sex, White self-identification, systolic blood pressure, previous CV disease, current smoking, BMI, HbA1c, eGFR, LDL cholesterol, and albuminuria, and (for the dulaglutide model) the interaction of dulaglutide and the change in biomarker. All interaction P values were >0.05. OR, odds ratio.

Relationship Between the 2-Year Change in Metabolite Biomarkers, Dulaglutide Assignment, and MACE

The relationship between ln (2-year/baseline) for each metabolite measured and dulaglutide assignment was analyzed using 135 separate adjusted linear regression models. Supplementary Table 6 lists dulaglutide’s β-coefficients along with the 95% CIs and P values for these metabolites. The two biomarkers for which the β-coefficient P values for the relationship to dulaglutide (Fig. 2B) were significant at a Bonferroni-corrected P value of 0.00037 (i.e., 0.05/135) were 2-hydroxybutyric acid (2-HB), for which dulaglutide assignment was associated with a greater reduction from baseline than placebo, and threonine, for which dulaglutide assignment was associated with a greater rise from baseline than placebo. Neither of these metabolites were significantly associated with MACE status (Table 2) at a significance threshold of 0.025 (i.e., 0.05/2 comparisons). Similar findings were noted in the sensitivity analysis (Supplementary Table 5).

In this nested case-control study and post hoc analysis of REWIND trial participants, dulaglutide was associated with lesser 2-year increases from baseline of NTproBNP and GDF15 levels versus placebo. Moreover, a smaller increase of these two proteins from baseline was associated with a lower incidence of MACE. Although dulaglutide was also associated with greater increases from baseline in C-peptide and threonine and a greater reduction in hsCRP and 2-HB versus placebo, these changes were not associated with MACE.

These new protein biomarker findings suggest that dulaglutide’s association with changes in NTproBNP and GDF15 are related to some of its CV benefits. The findings are consistent with the theoretical possibility that dulaglutide’s salutary effects on MACE in the REWIND trial (6) are mediated by a reduction in these biomarkers. However, the likeliest explanations are that dulaglutide reduces MACE and levels of these biomarkers through independent mechanisms, and/or that dulaglutide’s effect on processes leading to MACE is reflected in a fall in the level of these biomarkers.

NT-proBNP, the prohormone for the active natriuretic hormone B-type natriuretic peptide, is secreted in response to myocardial stretch (11). NT-proBNP is an established risk factor for incident CV outcomes (12,13), atrial fibrillation (14), and death (15), and its level rises in people with heart failure and falls with heart failure treatment (16). These observations and the salutary CV effects of natriuretic peptides (11) suggest that NT-proBNP either mitigates or reflects one or more processes underlying CV outcomes. Our finding that dulaglutide blunted the rise in NT-proBNP is consistent with reports that the GLP-1 RA liraglutide reduced this biomarker in people with heart failure, reduced ejection fraction heart failure (17), and in people with newly diagnosed type 2 diabetes (18). It is also consistent with a report that the GLP-1 RA lixisenatide reduced NT-proBNP more than placebo did in people with diabetes and recent acute coronary syndrome (19).

GDF15 is widely expressed in the liver, intestine, and kidneys. Its concentration rises in response to heart and kidney failure, cancers, inflammation, mitochondrial dysfunction, obesity, and metformin (20), and high levels are associated with reduced food intake and anti-inflammatory effects (21). GDF15 levels were also reduced in people assigned to dulaglutide in one small trial (8). Although elevated levels are independent risk factors for CV outcomes (22), accruing evidence suggests that GDF15 may have several CV benefits (21).

The metabolite findings of a dulaglutide-associated reduction in 2-HB has been previously reported (7) and is consistent with both the increase in threonine and an improved glucometabolic status (23). The absence of a significant relationship with MACE suggests either this change is unrelated to a CV benefit (24) or that the analysis was insufficiently powered to clearly detect such a benefit.

Strengths of these analyses include the careful matching of cases with controls, the large sample size, and the large number of MACE outcomes that were available for analyses. The stringent criteria for statistical significance and the use of a hypothesis-free discovery approach followed by a hypothesis-testing approach represents another strength that minimizes the likelihood of detecting spurious signals. Nevertheless, these analyses are limited by their retrospective nature, and the possibility that the biomarker changes associated with MACE may have been a consequence of MACE events that occurred prior to the 2-year follow-up biomarker level. The absence of a significant association between NT-proBNP changes and MACE in the sensitivity analysis may be due to this possibility. However, it may also be due to either a survival bias in the sensitivity analysis related to the exclusion of those most susceptible to MACE or the analysis of events that are more distant from, and therefore may be less related to, dulaglutide-associated change in this biomarker. These analyses are also limited by the number of protein biomarkers that were analyzed, the omission of troponin from the biomarker panel, and the fact that the analyses necessarily excluded people who could not provide a 2-year sample because of illness or death. Moreover, the stringent criteria for statistical significance precluded the identification of biomarkers that were highly correlated with those that were identified.

In summary, several GLP-1 RA drugs, including dulaglutide, reduce CV events. Evidence that reductions in two proteins that are associated with CVD, namely, NTproBNP and GDF15, are also associated with both dulaglutide and CV benefits highlights dulaglutide’s CV effects and supports further investigation into its underlying cardioprotective mechanisms.

Clinical trial reg. no. NCT01394952, clinicaltrials.gov

This article contains supplementary material online at https://doi.org/10.2337/figshare.22120097.

Funding. The REWIND trial and biomarker study was funded by Eli Lilly.

Duality of Interest. H.C.G. holds the McMaster-Sanofi Population Health Institute Chair in Diabetes Research and Care. He reports research grants from Eli Lilly, AstraZeneca, Merck, Novo Nordisk, and Sanofi; honoraria for speaking from Boehringer Ingelheim, Eli Lilly, Novo Nordisk, Sanofi, DKSH, Roche, and Zuellig; and consulting fees from Abbott, Covance, Eli Lilly, Novo Nordisk, Sanofi, Pfizer, Kowa, and Hanmi. M.A.B. is employed by Eli Lilly and Company and owns stock. H.M.C. reports research grants from Eli Lilly and Company, AstraZeneca, Regeneron, Pfizer, Roche, Sanofi, and Novo Nordisk; honoraria for speaking from Eli Lilly and Company and Regeneron; consulting fees from Eli Lilly and Company, Novartis, Regeneron, Sanofi, and Novo Nordisk; and shares in Bayer and Roche. A.H. is employed by Eli Lilly and Company and owns stock. M.L. owns shares in Eli Lilly. Y.L. is employed by Eli Lilly and Company and owns stock. V.P. is employed by Eli Lilly and Company and owns stock. H.-R.Q. is employed by Eli Lilly and Company and owns stock. G.R. is employed by Eli Lilly and Company and owns stock. L.R. reports research grants from the Swedish-Heart-Lung Foundation, Stockholm County Council, Erling Perssońs Foundation, Boehringer-Ingelheim, and Novo Nordisk, and honoraria for lecturing from Bayer AG, Novo Nordisk, Boehringer-Ingelheim, and Eli Lilly. J.M.W. is employed by Eli Lilly and Company and owns stock. K.L.D. is employed by Eli Lilly and Company and owns stock. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. H.C.G. wrote the first draft of the paper. S.-F.L., Y.L., and H.-R.Q. did the statistical analyses. V.P. performed the metabolomics analysis. J.M.W. analyzed the protein samples. All the authors researched the data and critically revised the paper. H.C.G. is the guarantor of the study and made the final decision to submit and publish the manuscript this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

1.
Nauck
MA
,
Quast
DR
,
Wefers
J
,
Pfeiffer
AFH
.
The evolving story of incretins (GIP and GLP-1) in metabolic and cardiovascular disease: a pathophysiological update
.
Diabetes Obes Metab
2021
;
23
(
Suppl. 3
):
5
29
2.
Lee
MMY
,
Kristensen
SL
,
Gerstein
HC
,
McMurray
JJV
,
Sattar
N
.
Cardiovascular and mortality outcomes with GLP-1 receptor agonists in patients with type 2 diabetes: a meta-analysis with the FREEDOM cardiovascular outcomes trial
.
Diabetes Metab Syndr
2022
;
16
:
102382
3.
Sattar
N
,
Lee
MMY
,
Kristensen
SL
, et al
.
Cardiovascular, mortality, and kidney outcomes with GLP-1 receptor agonists in patients with type 2 diabetes: a systematic review and meta-analysis of randomised trials
.
Lancet Diabetes Endocrinol
2021
;
9
:
653
662
4.
Buse
JB
,
Bain
SC
,
Mann
JFE
, et al.;
LEADER Trial Investigators
.
Cardiovascular risk reduction with liraglutide: an exploratory mediation analysis of the LEADER trial
.
Diabetes Care
2020
;
43
:
1546
1552
5.
Konig
M
,
Riddle
MC
,
Colhoun
HM
, et al
.
Exploring potential mediators of the cardiovascular benefit of dulaglutide in type 2 diabetes patients in REWIND
.
Cardiovasc Diabetol
2021
;
20
:
194
6.
Gerstein
HC
,
Colhoun
HM
,
Dagenais
GR
, et al.;
REWIND Investigators
.
Dulaglutide and cardiovascular outcomes in type 2 diabetes (REWIND): a double-blind, randomised placebo-controlled trial
.
Lancet
2019
;
394
:
121
130
7.
Pirro
V
,
Roth
KD
,
Lin
Y
, et al
.
Effects of tirzepatide, a dual GIP and GLP-1 RA, on lipid and metabolite profiles in subjects with type 2 diabetes
.
J Clin Endocrinol Metab
2022
;
107
:
363
378
8.
Wilson
JM
,
Lin
Y
,
Luo
MJ
, et al
.
The dual glucose-dependent insulinotropic polypeptide and glucagon-like peptide-1 receptor agonist tirzepatide improves cardiovascular risk biomarkers in patients with type 2 diabetes: a post hoc analysis
.
Diabetes Obes Metab
2022
;
24
:
148
153
9.
Gerstein
HC
,
Colhoun
HM
,
Dagenais
GR
, et al.;
REWIND Trial Investigators
.
Design and baseline characteristics of participants in the Researching cardiovascular Events with a Weekly INcretin in Diabetes (REWIND) trial on the cardiovascular effects of dulaglutide
.
Diabetes Obes Metab
2017
;
20
:
42
49
10.
Dagenais
GR
,
Rydén
L
,
Leiter
LA
, et al
.
Total cardiovascular or fatal events in people with type 2 diabetes and cardiovascular risk factors treated with dulaglutide in the REWIND trail: a post hoc analysis
.
Cardiovasc Diabetol
2020
;
19
:
199
11.
Kerkelä
R
,
Ulvila
J
,
Magga
J
.
Natriuretic peptides in the regulation of cardiovascular physiology and metabolic events
.
J Am Heart Assoc
2015
;
4
:
e002423
12.
Malachias
MVB
,
Jhund
PS
,
Claggett
BL
, et al
.
NT-proBNP by itself predicts death and cardiovascular events in high-risk patients with type 2 diabetes mellitus
.
J Am Heart Assoc
2020
;
9
:
e017462
13.
Sharma
A
,
Vaduganathan
M
,
Ferreira
JP
, et al
.
Clinical and biomarker predictors of expanded heart failure outcomes in patients with type 2 diabetes mellitus after a recent acute coronary syndrome: insights from the EXAMINE trial
.
J Am Heart Assoc
2020
;
9
:
e012797
14.
Staerk
L
,
Preis
SR
,
Lin
H
, et al
.
Protein biomarkers and risk of atrial fibrillation: the FHS
.
Circ Arrhythm Electrophysiol
2020
;
13
:
e007607
15.
Geng
Z
,
Huang
L
,
Song
M
,
Song
Y
.
N-terminal pro-brain natriuretic peptide and cardiovascular or all-cause mortality in the general population: a meta-analysis
.
Sci Rep
2017
;
7
:
41504
16.
Yancy
CW
,
Jessup
M
,
Bozkurt
B
, et al
.
2017 ACC/AHA/HFSA focused update of the 2013 ACCF/AHA guideline for the management of heart failure: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Failure Society of America
.
J Card Fail
2017
;
23
:
628
651
17.
Nielsen
R
,
Jorsal
A
,
Tougaard
RS
, et al
.
The impact of the glucagon-like peptide-1 receptor agonist liraglutide on natriuretic peptides in heart failure patients with reduced ejection fraction with and without type 2 diabetes
.
Diabetes Obes Metab
2020
;
22
:
2141
2150
18.
Lambadiari
V
,
Pavlidis
G
,
Kousathana
F
, et al
.
Effects of 6-month treatment with the glucagon like peptide-1 analogue liraglutide on arterial stiffness, left ventricular myocardial deformation and oxidative stress in subjects with newly diagnosed type 2 diabetes
.
Cardiovasc Diabetol
2018
;
17
:
8
19.
Wolsk
E
,
Claggett
B
,
Diaz
R
, et al
.
Increases in natriuretic peptides precede heart failure hospitalization in patients with a recent coronary event and type 2 diabetes mellitus
.
Circulation
2017
;
136
:
1560
1562
20.
Gerstein
HC
,
Pare
G
,
Hess
S
, et al.;
ORIGIN Investigators
.
Growth differentiation factor 15 as a novel biomarker for metformin
.
Diabetes Care
2017
;
40
:
280
283
21.
Wang
D
,
Day
EA
,
Townsend
LK
,
Djordjevic
D
,
Jørgensen
SB
,
Steinberg
GR
.
GDF15: emerging biology and therapeutic applications for obesity and cardiometabolic disease
.
Nat Rev Endocrinol
2021
;
17
:
592
607
22.
Gerstein
HC
,
Paré
G
,
McQueen
MJ
, et al.;
Outcome Reduction With Initial Glargine Intervention Trial Investigators
.
Identifying novel biomarkers for cardiovascular events or death in people with dysglycemia
.
Circulation
2015
;
132
:
2297
2304
23.
Cobb
J
,
Eckhart
A
,
Motsinger-Reif
A
,
Carr
B
,
Groop
L
,
Ferrannini
E
.
α-Hydroxybutyric acid is a selective metabolite biomarker of impaired glucose tolerance
.
Diabetes Care
2016
;
39
:
988
995
24.
Ferrannini
E
.
Tirzepatide as an insulin sensitizer
.
J Clin Endocrinol Metab
2022
;
107
:
e1752
e1753
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/journals/pages/license.