Did women benefit more than men in REWIND?

The point estimate was numerically lower in women (HR 0.85) than men (HR 0.90), but the confidence intervals overlapped and the interaction p-value was 0.55. There is no statistical evidence of a sex-based difference in treatment effect.

Does baseline BMI affect dulaglutide's cardiovascular benefit?

No. Patients with BMI below and above 32 kg/m² showed nearly identical hazard ratios (0.86 vs. 0.90, p-interaction 0.63). The benefit does not appear mediated solely through weight loss.

Was the MACE reduction significant in patients without prior heart disease?

The HR in the primary prevention subgroup was 0.87 (95% CI 0.74 to 1.02). While the confidence interval crossed 1.0 for that individual stratum, the p-interaction of 0.97 confirmed no heterogeneity versus the secondary prevention group. The overall trial result (HR 0.88, p = 0.026) applies to both populations.

How does REWIND's primary prevention signal compare to LEADER?

LEADER enrolled 81% of participants with established CVD. Its primary prevention subgroup showed a non-significant HR of 0.95. REWIND, with 69% primary prevention enrollment, produced a more balanced and consistent signal across both strata.

Did patients with kidney disease benefit from dulaglutide in REWIND?

Participants with eGFR <60 mL/min/1.73 m² had a consistent point estimate (HR 0.89). A separate renal sub-study confirmed a significant reduction in the composite renal endpoint (HR 0.85, 95% CI 0.77 to 0.93).

Does baseline HbA1c affect the cardiovascular benefit of dulaglutide?

No significant interaction was found (p = 0.55). The numerically larger benefit in the ≥7.2% group (HR 0.86 vs. 0.91) was not statistically distinct, and dulaglutide's cardiovascular indication is not restricted by glycemic control.

How long was REWIND's follow-up compared to other GLP-1 RA cardiovascular trials?

REWIND's median follow-up was 5.4 years, the longest of any completed GLP-1 RA CVOT. LEADER followed patients for 3.8 years and SUSTAIN-6 for 2.1 years. The longer duration gave subgroup analyses more statistical power.

Did older patients tolerate dulaglutide well enough to benefit?

Patients aged 66 and older showed an HR of 0.90, consistent with the younger subgroup (HR 0.86, p-interaction 0.65). The benefit was not attenuated by age.

Should dulaglutide be preferred over semaglutide for primary prevention patients?

REWIND provides the strongest GLP-1 RA evidence in primary prevention. SELECT tested semaglutide 2.4 mg in a different population (obesity with established CVD, not primary prevention T2D). Direct head-to-head data for primary prevention do not exist. The choice depends on individual patient factors, cost, and formulary coverage.

REWIND Subgroup Analyses: Who Responded Most and Least to Dulaglutide

Q: Were there racial or ethnic differences in treatment response?

The primary REWIND publication did not report hazard ratios stratified by race or ethnicity. The enrolled population was approximately 76% White, limiting conclusions about treatment effect heterogeneity in underrepresented groups.

Q: Does baseline HbA1c affect the cardiovascular benefit of dulaglutide?

No significant interaction was found (p = 0.55). The numerically larger benefit in the ≥7.2% group (HR 0.86 vs. 0.91) was not statistically distinct, and dulaglutide's cardiovascular indication is not restricted by glycemic control.

Q: How long was REWIND's follow-up compared to other GLP-1 RA cardiovascular trials?

REWIND's median follow-up was 5.4 years, the longest of any completed GLP-1 RA CVOT. LEADER followed patients for 3.8 years and SUSTAIN-6 for 2.1 years. The longer duration gave subgroup analyses more statistical power.

Q: Did older patients tolerate dulaglutide well enough to benefit?

Patients aged 66 and older showed an HR of 0.90, consistent with the younger subgroup (HR 0.86, p-interaction 0.65). The benefit was not attenuated by age.

Q: Should dulaglutide be preferred over semaglutide for primary prevention patients?

REWIND provides the strongest GLP-1 RA evidence in primary prevention. SELECT tested semaglutide 2.4 mg in a different population (obesity with established CVD, not primary prevention T2D). Direct head-to-head data for primary prevention do not exist. The choice depends on individual patient factors, cost, and formulary coverage.

By HealthRX.com Medical Team

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

Clinical image for REWIND Subgroup Analyses: Who Responded Most and Least to Dulaglutide 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 | | 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 to 0.99; p = 0.026) |

Why the Subgroup Data Matters More Than Usual

Most GLP-1 receptor agonist cardiovascular outcomes trials enrolled populations with established atherosclerotic cardiovascular disease (ASCVD). REWIND was different. Roughly 69% of participants had cardiovascular risk factors only, without a prior event. That enrollment profile made the subgroup analyses unusually important: they could answer whether dulaglutide's benefit was confined to secondary prevention (where the signal is easiest to detect) or extended to the larger, harder-to-treat primary prevention population.

The trial's 5.4-year median follow-up, the longest of any completed GLP-1 RA CVOT at the time, also gave the subgroup comparisons more statistical power than shorter studies could provide.

Pre-Specified Subgroup Framework

The REWIND investigators pre-specified 14 subgroup comparisons in the statistical analysis plan. We organize them below into three tiers based on clinical prescribing relevance, not the order they appeared in the supplement.

Tier 1: Patient-selection subgroups (directly shape who gets prescribed dulaglutide for CV risk)

Prior cardiovascular event (yes vs. no)
Age (<66 vs. ≥66 years)
Sex (male vs. female)
HbA1c (<7.2% vs. ≥7.2%)

Tier 2: Physiologic-modifier subgroups (affect expected magnitude of benefit)

BMI (<32 vs. ≥32 kg/m²)
Duration of diabetes (<5 vs. 5 to <10 vs. ≥10 years)
eGFR (<60 vs. ≥60 mL/min/1.73 m²)
Baseline statin use (yes vs. no)

Tier 3: Regional and demographic subgroups (inform generalizability)

Geographic region (Americas, Europe, Asia/Pacific)
Race/ethnicity (where reported)

All subgroup analyses used Cox proportional-hazards models with interaction terms. A p-interaction <0.05 threshold flagged heterogeneity, though the investigators acknowledged the analyses were not powered for individual subgroup significance.

Tier 1 Results: The Clinically Decisive Subgroups

Prior CVD Status

This was the most scrutinized split. Among participants with prior cardiovascular disease (n = 3,114), the hazard ratio was 0.87 (95% CI 0.74 to 1.02). Among those without a prior event (n = 6,787), the HR was 0.87 (95% CI 0.74 to 1.02). The p-interaction was 0.97, showing no heterogeneity between groups.

That near-identical point estimate across both strata was clinically notable. LEADER (liraglutide) and SUSTAIN-6 (semaglutide) had enrolled populations where 73% and 83% of participants, respectively, had established CVD. REWIND's result suggested dulaglutide's MACE reduction was not dependent on the presence of existing atherosclerotic plaque burden.

The 2019 ADA Standards of Care subsequently cited REWIND when expanding GLP-1 RA recommendations beyond patients with established ASCVD to include those with multiple cardiovascular risk factors.

Age

Participants were split at the median age of 66 years.

| Age group | n | HR (95% CI) | |---|---|---| | <66 years | 4,949 | 0.86 (0.73 to 1.02) | | ≥66 years | 4,952 | 0.90 (0.78 to 1.04) |

p-interaction = 0.65. The hazard ratios tracked closely. Older adults, who often face dose-limiting GI side effects with GLP-1 RAs, appeared to derive comparable benefit.

Sex

Women made up 46% of the REWIND population, a substantially higher proportion than LEADER (36%) or SUSTAIN-6 (39%).

| Sex | n | HR (95% CI) | |---|---|---| | Male | 5,345 | 0.90 (0.79 to 1.04) | | Female | 4,556 | 0.85 (0.71 to 1.01) |

p-interaction = 0.55. The numerically lower point estimate in women was not statistically distinct from the male subgroup. A post-hoc pooled analysis of GLP-1 RA CVOTs later confirmed that sex does not meaningfully modify the class-level cardiovascular benefit.

Baseline HbA1c

Participants were split at the cohort median HbA1c of 7.2%.

| HbA1c | n | HR (95% CI) | |---|---|---| | <7.2% | 4,644 | 0.91 (0.77 to 1.07) | | ≥7.2% | 5,227 | 0.86 (0.74 to 0.99) |

p-interaction = 0.55. The benefit trended numerically larger in the higher-HbA1c group, but the confidence intervals overlapped broadly. This pattern aligns with dulaglutide's FDA prescribing information, which does not restrict its cardiovascular indication by glycemic control status.

Tier 2 Results: Physiologic Modifiers

BMI

| BMI category | n | HR (95% CI) | |---|---|---| | <32 kg/m² | 4,891 | 0.86 (0.74 to 0.99) | | ≥32 kg/m² | 5,010 | 0.90 (0.77 to 1.06) |

p-interaction = 0.63. There was no evidence that higher baseline weight blunted the cardiovascular benefit. This mattered because clinicians sometimes assume GLP-1 RA effects are mediated primarily through weight loss; the REWIND subgroup data supports a benefit independent of baseline adiposity.

Diabetes Duration

The trial split duration into three categories:

| Duration | HR (95% CI) | |---|---| | <5 years | 0.86 (0.68 to 1.08) | | 5 to <10 years | 0.87 (0.71 to 1.07) | | ≥10 years | 0.89 (0.76 to 1.04) |

p-interaction = 0.94. The flat interaction is reassuring for prescribers managing newly diagnosed patients alongside those with long-standing disease.

Baseline Renal Function

| eGFR | n | HR (95% CI) | |---|---|---| | ≥60 mL/min/1.73 m² | 7,558 | 0.87 (0.77 to 0.99) | | <60 mL/min/1.73 m² | 2,199 | 0.89 (0.73 to 1.08) |

p-interaction = 0.78. Patients with moderate CKD showed a consistent point estimate. A dedicated renal sub-study of REWIND published in The Lancet Diabetes & Endocrinology in 2019 confirmed a separate composite renal endpoint benefit (HR 0.85, 95% CI 0.77 to 0.93), which strengthened the case for dulaglutide use in diabetic kidney disease.

Statin Use at Baseline

Roughly 66% of participants were on statins at enrollment.

| Statin use | HR (95% CI) | |---|---| | Yes | 0.87 (0.76 to 0.99) | | No | 0.91 (0.75 to 1.09) |

p-interaction = 0.68. Dulaglutide's benefit appeared additive to background lipid-lowering therapy, not dependent on it.

Tier 3: Geographic Region and Race/Ethnicity

REWIND enrolled across 24 countries. Regional hazard ratios:

| Region | HR (95% CI) | |---|---| | Americas | 0.88 (0.75 to 1.03) | | Europe | 0.88 (0.74 to 1.06) | | Asia/Pacific | 0.86 (0.65 to 1.13) |

p-interaction = 0.98. The consistency across geographically and ethnically diverse populations supports external validity.

Granular race/ethnicity breakdowns were limited in the primary publication. The enrolled population was approximately 76% White, 7% Black, 9% Hispanic/Latino, and 4% East Asian. The trial was not powered to detect treatment-by-race interactions, and the investigators did not report separate hazard ratios stratified by self-reported race in the main paper. This remains a gap. Post-hoc analyses have not been published at sufficient granularity to clarify whether the benefit is uniform across racial subgroups.

What the Forest Plot Actually Shows

Across all 14 pre-specified subgroups, every point estimate fell between 0.83 and 0.93. No interaction p-value crossed the 0.05 threshold. The published forest plot is unusually uniform compared to other GLP-1 RA CVOTs, where subgroups with <60% prior CVD often showed attenuated or null effects.

The practical interpretation: dulaglutide's 12% MACE reduction is not concentrated in a single phenotype. Clinicians selecting a GLP-1 RA for cardiovascular risk reduction do not need to restrict consideration to young, male, high-BMI, or secondary-prevention patients.

Limitations the Authors Acknowledged

Multiplicity. Fourteen subgroup tests inflate the false-negative and false-positive rate. The authors applied no formal correction for multiple comparisons.
Power. Individual subgroups were underpowered to detect treatment effects. The confidence intervals for smaller strata (Asia/Pacific, eGFR <60) were wide.
Single dose. REWIND tested only dulaglutide 1.5 mg weekly. Whether lower doses (0.75 mg) would produce consistent subgroup patterns is unknown.
Race/ethnicity reporting. The trial lacked the granularity to draw conclusions about treatment effect heterogeneity by race. Black and Hispanic/Latino participants were underrepresented relative to the U.S. diabetes population.
Post-hoc analyses. Some subgroup splits (e.g., by baseline NT-proBNP or hs-CRP) were exploratory and should be interpreted with appropriate caution.

Clinical Translation

The 2023 ADA/EASD consensus report cites REWIND's primary prevention signal as one reason GLP-1 RAs are now recommended for patients with type 2 diabetes and multiple cardiovascular risk factors, not only those with established ASCVD. The subgroup consistency is part of the evidence base supporting that expanded indication.

For prescribers weighing dulaglutide against other GLP-1 RAs: SELECT (semaglutide 2.4 mg) later demonstrated a larger MACE reduction in patients with obesity and established CVD, but did not enroll a comparable primary prevention fraction. REWIND remains the strongest evidence for GLP-1 RA cardiovascular benefit in the primary prevention population as of this review date.

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. Dulaglutide and renal outcomes in type 2 diabetes: an exploratory analysis of the REWIND randomised, placebo-controlled trial. Lancet Diabetes Endocrinol. 2019;7(12):927-936. PubMed
American Diabetes Association. Standards of Medical Care in Diabetes, 2019. Diabetes Care. 2019;42(Suppl 1). PubMed
Dulaglutide (Trulicity) prescribing information. U.S. Food and Drug Administration. FDA Label
Davies MJ, Aroda VR, Collins BS, et al. Management of hyperglycaemia in type 2 diabetes, 2022. A consensus report by the ADA and EASD. Diabetologia. 2022;65(12):1925-1966. PubMed
Giugliano D, Scappaticcio L, Longo M, et al. GLP-1 receptor agonists and cardiorenal outcomes in type 2 diabetes: an updated meta-analysis of eight CVOTs. Cardiovasc Diabetol. 2021;20(1):189. PubMed