What does REWIND stand for?

Researching Cardiovascular Events with a Weekly Incretin in Diabetes. The trial tested dulaglutide 1.5 mg weekly against placebo in 9,901 adults with type 2 diabetes and cardiovascular risk factors.

Did REWIND include patients without prior heart disease?

Yes. Approximately 69% of participants had no prior cardiovascular events at enrollment. This makes REWIND the most primary-prevention-weighted GLP-1 RA CVOT completed to date, per the trial design.

How long did the REWIND trial last?

Median follow-up was 5.4 years, driven by an event-based design requiring 1,200 primary-endpoint events. This is substantially longer than most other GLP-1 CVOTs.

Did dulaglutide reduce stroke risk in REWIND?

Yes. Non-fatal stroke was reduced by 24% (HR 0.76, 95% CI 0.61, 0.95), the most pronounced individual component benefit. A dedicated sub-analysis confirmed this signal across ischemic stroke subtypes.

Was REWIND a superiority or non-inferiority trial?

Superiority. The trial was powered from inception to detect a true MACE reduction, not merely to rule out excess cardiovascular harm, as described in the protocol.

How does REWIND compare to LEADER and SUSTAIN-6?

REWIND enrolled a lower-risk, more primary-prevention-oriented population and followed patients nearly twice as long as SUSTAIN-6. All three trials showed MACE reduction with GLP-1 RAs, but direct cross-trial comparisons are unreliable due to differing populations and designs.

Did the FDA update dulaglutide's label based on REWIND?

Yes. The current Trulicity prescribing information includes an indication for reducing the risk of MACE in adults with type 2 diabetes and established cardiovascular disease or multiple cardiovascular risk factors.

What was the discontinuation rate in REWIND?

Approximately 24% in each arm stopped study drug before trial end. The primary ITT analysis included all randomized participants regardless, providing a conservative ("treatment policy") estimate of benefit.

Does REWIND prove dulaglutide works for primary CV prevention?

REWIND showed a consistent MACE reduction across patients with and without prior CV events, but the subgroup analysis was not independently powered. The primary publication supports benefit across the full risk spectrum enrolled, though dedicated primary-prevention trials would strengthen this conclusion.

Inside the REWIND Methodology: What Most Summaries Skip

Q: What was the primary endpoint in REWIND?

Three-point MACE: a composite of cardiovascular death, non-fatal myocardial infarction, and non-fatal stroke, adjudicated by a blinded committee.

By HealthRX.com Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

Clinical image for Inside the REWIND Methodology: What Most Summaries Skip Image: HealthRX.com clinical image

At a glance

| Parameter | Detail | |---|---| | Trial name | REWIND (Researching Cardiovascular Events with a Weekly Incretin in Diabetes) | | N | 9,901 | | Intervention | Dulaglutide 1.5 mg subcutaneous, once weekly | | Comparator | Matched placebo, once weekly | | Median follow-up | 5.4 years | | Primary endpoint | First occurrence of 3-point MACE (CV death, non-fatal MI, non-fatal stroke) | | Key result | HR 0.88 (95% CI 0.79, 0.99; p = 0.026) | | Registry | NCT01394952 |

Why the Population Matters More Than the Hazard Ratio

Most GLP-1 receptor agonist cardiovascular outcome trials (CVOTs) enrolled patients with established atherosclerotic cardiovascular disease (ASCVD). REWIND took a different path. Per the published protocol, eligibility required age 50 or older with established CV disease or age 55 or older with at least one additional risk factor: coronary, cerebrovascular, or peripheral artery stenosis; left ventricular hypertrophy; an eGFR <60 mL/min/1.73 m²; or albuminuria. This meant roughly 69% of randomized participants did not have prior CV events at baseline, a primary-prevention-heavy cohort that is rare among completed CVOTs.

That design choice has a direct clinical consequence. When the primary results paper reported a 12% relative risk reduction in three-point MACE, the finding applied to a population that looks far more like everyday endocrinology practice than the high-risk cohorts of LEADER (81% established ASCVD) or SUSTAIN-6 (83% established ASCVD). Subgroup analysis in the primary publication showed consistent benefit regardless of prior CV event history, though confidence intervals were wider in the secondary-prevention subgroup simply because of fewer events.

Randomization and Blinding Architecture

REWIND used a centralized, computer-generated randomization scheme stratified by site and prior CV event status. Allocation was 1:1. Both dulaglutide and placebo were supplied in identical pre-filled pens to maintain double-blinding. Investigators, participants, and the adjudication committee were all masked.

A feature worth noting: the trial design paper specified that background diabetes therapy could be adjusted at the discretion of the treating physician during the trial. This pragmatic approach improved generalizability but introduced a confounder. Over the median 5.4-year follow-up, between-group HbA1c differences narrowed from an initial 0.6% separation to roughly 0.2% by trial end. The CV benefit therefore cannot be fully explained by glucose lowering alone, a pattern consistent with other GLP-1 RA CVOTs.

The Primary Endpoint: Three-Point MACE, Defined Precisely

REWIND used a standard FDA-endorsed three-point MACE composite: cardiovascular death, non-fatal myocardial infarction, and non-fatal stroke. Each component was adjudicated by an independent clinical events committee blinded to treatment assignment.

The breakdown across individual components, as reported in the primary results:

| Component | Dulaglutide (n/N) | Placebo (n/N) | HR (95% CI) | |---|---|---|---| | Composite MACE | 594 / 4,949 | 663 / 4,952 | 0.88 (0.79, 0.99) | | CV death | 317 | 346 | 0.91 (0.78, 1.06) | | Non-fatal MI | 205 | 212 | 0.96 (0.79, 1.16) | | Non-fatal stroke | 135 | 175 | 0.76 (0.61, 0.95) |

The composite reached significance. Individual components did not, except non-fatal stroke, which showed a striking 24% relative reduction. A dedicated stroke sub-analysis later confirmed this signal was consistent across ischemic stroke subtypes, making REWIND one of the few diabetes CVOTs to show a clear cerebrovascular benefit.

Statistical Design: Event-Driven, Not Time-Driven

REWIND was powered for superiority from the start, not merely non-inferiority. The statistical plan required at least 1,200 adjudicated primary-endpoint events to provide 90% power to detect a hazard ratio of 0.87 or lower at a two-sided alpha of 0.05. This event-driven design meant the trial continued until the event threshold was met, which extended median follow-up to 5.4 years, considerably longer than LEADER (3.8 years) or SUSTAIN-6 (2.1 years).

The primary analysis used a Cox proportional hazards model stratified by the same factors used at randomization (site cluster and prior CV event status). No interim analyses for efficacy were pre-specified, though a single interim futility analysis was planned and conducted. The Data Safety Monitoring Board reviewed unblinded data periodically for safety. The protocol paper describes this framework in detail.

The intention-to-treat principle governed the primary analysis: all randomized patients were analyzed in their assigned groups regardless of treatment discontinuation. Roughly 24% of participants in each arm stopped study drug before the trial ended, a notable discontinuation rate that could have diluted the treatment effect. A per-protocol sensitivity analysis was not the primary estimand, but the consistency of results across pre-specified sensitivity analyses strengthened confidence in the ITT finding.

The Estimand Question: What Was Actually Being Measured?

REWIND was designed before the ICH E9(R1) addendum formalized estimand frameworks, but the trial's structure lends itself to retrospective estimand analysis. The ITT approach used in REWIND estimates a "treatment policy" strategy: the effect of being assigned to dulaglutide, regardless of whether the patient stayed on drug. Given the ~24% discontinuation rate, this is a conservative estimate. A hypothetical "on-treatment" estimand (what would have happened if everyone stayed on drug) would likely show a larger treatment effect.

This matters for clinical interpretation. The HR of 0.88 is what a health system can expect when it prescribes dulaglutide to a REWIND-like population, knowing that real-world adherence will be imperfect. The biological efficacy of the molecule may be larger than 12%, but the trial was pragmatically designed to answer the question clinicians actually face.

Comparator Choice and Background Therapy

The comparator was placebo, not an active GLP-1 RA or another glucose-lowering agent. Per the published design, investigators were encouraged to optimize glycemic control, blood pressure, and lipid management in both arms according to local guidelines. This means REWIND tested dulaglutide on top of standard care, not instead of alternative treatments.

Statin use exceeded 66% in both groups at baseline. ACE inhibitors or ARBs were used in over 81% of participants. These high rates of guideline-directed medical therapy meant the incremental CV benefit attributed to dulaglutide was achieved against a strong background of already-optimized care, a point the 2019 ESC/EASD guidelines cited when recommending GLP-1 RAs in patients with T2D and CV risk.

What REWIND Did Not Answer

No trial answers every question. REWIND's acknowledged limitations include:

Dose ceiling. Only dulaglutide 1.5 mg was tested. Higher doses (3.0 mg and 4.5 mg, now approved per the Trulicity prescribing information) were not available during the trial period. Whether higher doses provide additional CV protection is unknown.
No head-to-head GLP-1 RA comparison. REWIND cannot tell us whether dulaglutide is superior, equivalent, or inferior to semaglutide or liraglutide for CV risk reduction. Cross-trial comparisons with LEADER or SUSTAIN-6 are hypothesis-generating only.
Ethnic diversity. The cohort was predominantly white (~76%), with limited representation from populations at disproportionate cardiometabolic risk.
HbA1c convergence. The narrowing HbA1c gap between arms over time raises the question of whether MACE reduction was glucose-mediated, pleiotropic, or both. REWIND was not designed to isolate the mechanism.
Discontinuation rate. At ~24% per arm, the treatment policy estimand is conservative. This does not invalidate the finding, but it does complicate biological interpretation.

How REWIND Changed Practice

Following the 2019 primary publication, the ADA Standards of Care updated recommendations to position GLP-1 RAs with proven CV benefit (including dulaglutide) as preferred agents in patients with T2D and established ASCVD or high CV risk, independent of baseline HbA1c. The FDA subsequently updated the Trulicity label to include a CV risk reduction indication.

REWIND's unique contribution was extending the evidence base beyond secondary prevention. Earlier CVOTs, including EXSCEL (exenatide, which failed to show superiority for MACE), focused almost entirely on patients with prior events. REWIND demonstrated that CV protection with a GLP-1 RA can begin before the first heart attack or stroke, not just after.

Frequently asked questions

›

References

Gerstein HC, Colhoun HM, Dagenais GR, et al. Dulaglutide and cardiovascular outcomes in type 2 diabetes (REWIND): a double-blind, randomised placebo-controlled trial. Lancet. 2019;394(10193):121-130. PubMed
Gerstein HC, Colhoun HM, Dagenais GR, et al. 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. 2018;20(1):42-49. PubMed
Marso SP, Daniels GH, Poulter NR, et al. Liraglutide and cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2016;375(4):311-322. PubMed
Marso SP, Bain SC, Consoli A, et al. Semaglutide and cardiovascular outcomes in patients with type 2 diabetes. N Engl J Med. 2016;375(19):1834-1844. PubMed
Trulicity (dulaglutide) prescribing information. Eli Lilly. Revised 2022. FDA Label
Cosentino F, Grant PJ, Aboyans V, et al. 2019 ESC Guidelines on diabetes, pre-diabetes, and cardiovascular diseases. Eur Heart J. 2020;41(2):255-323. PubMed