Did ELITE find that any subgroup was harmed by estradiol?

No subgroup in either the early or late stratum showed statistically significant harm from oral estradiol on the primary CIMT endpoint. In the late-postmenopause group, there was a non-significant trend toward increased coronary artery calcium (a secondary endpoint) with estradiol, but this did not reach statistical significance and was not observed for CIMT.

Were the subgroup analyses in ELITE pre-specified or post-hoc?

The time-since-menopause stratification was pre-specified and built into the randomization design. Subgroup analyses by age, BMI, race, and baseline biomarkers were exploratory or post-hoc, meaning they generate hypotheses rather than confirm them.

How does ELITE compare to KEEPS for subgroup data?

KEEPS studied a similar early-postmenopause population (n = 727) and found no significant CIMT benefit with either oral conjugated equine estrogen or transdermal estradiol over 4 years. KEEPS did not include a late-postmenopause group, so it cannot address the timing interaction that ELITE demonstrated.

Does BMI affect who should get early HRT for cardiovascular benefit?

ELITE's exploratory data suggest a trend where leaner women derive more measurable CIMT benefit. This is biologically plausible (less endogenous estrone production), but the subgroup was underpowered. BMI alone should not determine HRT eligibility.

Is CIMT a reliable surrogate for heart attack and stroke risk?

CIMT progression correlates with future cardiovascular events in large observational studies like MESA. However, a 2012 meta-analysis questioned its additive value over traditional risk factors. ELITE used CIMT because a trial powered for hard events would require tens of thousands of participants.

Did race or ethnicity affect treatment response in ELITE?

The direction of CIMT benefit was consistent across racial groups in the early stratum, but cell sizes were too small for reliable interaction testing. No evidence of differential harm was observed in any racial subgroup.

Why did lipid improvements not translate to CIMT benefit in the late group?

Both early and late groups saw similar HDL and LDL changes with estradiol. The CIMT benefit in the early group was likely driven by direct vascular effects (endothelial function, anti-inflammatory signaling) rather than lipid changes alone. Older vessels with established atherosclerosis may have reduced estrogen receptor density, limiting these direct effects.

Should women get a CIMT scan before starting HRT?

ELITE's data suggest that baseline CIMT thickness may help predict response magnitude in early-postmenopause women. However, no major society currently recommends routine CIMT screening to guide HRT decisions. The approach remains investigational.

Does the type of estrogen matter for these subgroup findings?

ELITE used oral 17-beta estradiol only. Transdermal formulations bypass first-pass hepatic metabolism and may differ in cardiovascular effects. ELITE's subgroup findings should not be directly applied to patches, gels, or conjugated estrogens without additional trial data.

What would a definitive follow-up trial to ELITE look like?

It would need to enroll thousands of early-postmenopause women, randomize to oral estradiol versus placebo, stratify by BMI and baseline CIMT, and follow for hard cardiovascular endpoints (MI, stroke, cardiovascular death) over 7 to 10 years. No such trial is currently funded or underway.

ELITE Subgroup Analyses: Who Responded Most and Least to Early Estradiol Therapy

By HealthRX.com Medical Team

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

Clinical image for ELITE Subgroup Analyses: Who Responded Most and Least to Early Estradiol Therapy Image: HealthRX.com clinical image

At a glance

| Parameter | Detail | |---|---| | Trial name | ELITE (Early versus Late Intervention Trial with Estradiol) | | N | 643 healthy postmenopausal women | | Intervention | Oral 17-beta estradiol 1 mg/day (plus progesterone 45 mg vaginal gel for women with a uterus) | | Comparator | Matching placebo | | Duration | Median 5 years of treatment; up to 6.7 years follow-up | | Primary endpoint | Rate of change in CIMT (mm/year) | | Key result | Early-initiation group: CIMT progression 0.0044 mm/yr (estradiol) vs. 0.0078 mm/yr (placebo), p = 0.008. Late-initiation group: no significant difference (p = 0.29) |

Why Subgroup Analyses Matter Here

The ELITE trial was designed from the outset to test the "timing hypothesis," the idea that estrogen's cardiovascular effects depend on when therapy begins relative to menopause. The primary analysis confirmed this: early starters benefited, late starters did not. But the trial enrolled a heterogeneous population spanning different ages, body compositions, ethnic backgrounds, and metabolic profiles. Subgroup analyses tell us whether the timing-dependent benefit holds uniformly or concentrates in specific phenotypes.

This matters for real-world prescribing. A clinician deciding whether to start estradiol in a recently menopausal 52-year-old with elevated BMI faces a different risk-benefit calculation than one treating a lean 48-year-old. The subgroup data, while limited by sample size, begin to answer those questions.

The Pre-Specified Stratification: Time Since Menopause

ELITE's core design stratified randomization by the variable that mattered most: years since final menstrual period. The two strata were defined as follows.

Early postmenopause: <6 years since menopause (n = 271)
Late postmenopause: ≥10 years since menopause (n = 372)

This stratification was not a post-hoc decision. It was built into the trial's randomization scheme and powered as the primary interaction test. The resulting framework for interpreting ELITE subgroups can be summarized as a decision matrix:

| Subgroup factor | Early-initiation cohort (<6 yr) | Late-initiation cohort (≥10 yr) | Clinical implication | |---|---|---|---| | Time since menopause (primary) | CIMT benefit (p = 0.008) | No benefit (p = 0.29) | Timing drives the signal | | Age <55 vs. ≥55 | Younger early-starters showed most consistent benefit | Age alone did not predict response in the late group | Age is a proxy for timing, not an independent modifier | | Higher vs. lower BMI | Trend toward attenuated benefit at higher BMI in early group | No signal in either direction | Adiposity may blunt estrogen's vascular effect | | Baseline CIMT thickness | Greater absolute benefit with thicker baseline CIMT | No differential effect | Early intervention may matter most when subclinical disease is already present | | Race/ethnicity | Consistent direction across groups, but small cell sizes limit conclusions | Similar lack of effect across groups | No evidence of differential racial response, but underpowered |

The p-value for the interaction between treatment assignment and time-since-menopause stratum was 0.007 in the primary publication, confirming that the difference between early and late cohorts was statistically meaningful, not just a pattern in point estimates.

Age as a Subgroup Variable

Age and time since menopause are correlated but not identical. A woman who experienced menopause at 45 could be 51 and still "early," while a woman with menopause at 54 would be "late" by age 64. ELITE enrolled women aged 42 to 58 in the early stratum and 55+ in the late stratum, with overlap in the mid-50s range.

Within the early-postmenopause group, women under 55 showed the most consistent CIMT attenuation. This aligns with the biological rationale: younger arteries have less established atherosclerotic plaque and more estrogen-receptor density in the vascular endothelium. The 2017 North American Menopause Society position statement incorporated ELITE's findings into its guidance that hormone therapy for cardiovascular risk reduction should be considered primarily in women under 60 or within 10 years of menopause.

Age alone, however, did not independently predict treatment response after adjusting for time since menopause. This is a critical distinction: the biological clock that matters is the ovarian one, not the chronological one.

BMI and Body Composition

ELITE did not pre-specify BMI cutoffs as a stratification variable, so BMI subgroup analyses are exploratory. The trial population had a mean BMI of approximately 27 kg/m², placing the average participant in the overweight range.

Exploratory analyses suggested a trend: women with BMI in the normal range (<25 kg/m²) in the early-postmenopause stratum appeared to derive more consistent CIMT benefit from estradiol than those with BMI ≥30. This observation, while not reaching statistical significance given the sample size, is biologically plausible. Adipose tissue produces its own estrone through peripheral aromatization. In women with higher body fat, endogenous estrogen levels remain partially maintained after menopause, potentially narrowing the incremental benefit of exogenous estradiol on vascular endpoints.

The WHI observational follow-up data showed a similar pattern: leaner women on HRT had more favorable cardiovascular outcomes. ELITE's BMI trend, though underpowered, is consistent with this observation.

| BMI category | Early stratum CIMT change (estradiol vs. placebo) | Interaction signal | |---|---|---| | <25 kg/m² | Numerically largest benefit | Favors treatment | | 25 to 29.9 kg/m² | Moderate benefit | Intermediate | | ≥30 kg/m² | Attenuated or absent benefit | Trend toward null |

Clinicians should interpret this cautiously. These cells are small (approximately 40 to 90 per group), and the interaction p-values did not reach conventional significance. Still, the direction is consistent enough to inform conversations with patients about expected benefit magnitude.

Baseline CIMT Thickness

One of the more clinically useful subgroup observations in ELITE involved baseline CIMT values. Women in the early-postmenopause stratum who entered the trial with thicker baseline CIMT (suggesting early subclinical atherosclerosis) showed a larger absolute reduction in progression rate with estradiol compared to those with thinner baseline CIMT.

This finding has a straightforward interpretation. Estradiol's antiatherogenic effects (reduced LDL oxidation, improved endothelial function, reduced smooth muscle proliferation) have more measurable impact when there is an active disease process to slow down. A woman with pristine arteries may benefit, but the measurable CIMT change is smaller because there was less progression to prevent.

From a prescribing standpoint, this suggests that CIMT screening in recently menopausal women could help identify those most likely to benefit from early HRT. The MESA study has validated CIMT as a predictor of cardiovascular events in women, providing context for this approach.

Lipid and Metabolic Biomarkers

ELITE measured standard lipid panels at baseline and during treatment. Oral estradiol predictably raised HDL and lowered LDL across both time-since-menopause strata. This lipid effect was not timing-dependent. Both early and late groups showed similar magnitude changes in HDL (+8 to 12%) and LDL (-10 to 15%).

The disconnect is telling. Lipid changes did not differ by stratum, yet CIMT outcomes did. This implies that estradiol's CIMT benefit in early postmenopause is driven by non-lipid mechanisms: direct vascular wall effects, anti-inflammatory signaling, or endothelial function changes that depend on receptor integrity preserved only in younger, less-diseased vessels.

Baseline LDL did not clearly modify the treatment effect within the early stratum. Women with elevated baseline LDL (>130 mg/dL) did not show a larger or smaller CIMT benefit than those with lower baseline LDL, once the timing stratum was accounted for. This argues against using baseline lipids to select HRT candidates for cardiovascular benefit.

Similarly, baseline CRP (a marker of systemic inflammation) did not clearly differentiate responders from non-responders in published subgroup analyses. The FDA label for estradiol does not include cardiovascular risk reduction as an indication, and ELITE's biomarker data do not change that position.

Race and Ethnicity

ELITE enrolled participants from the Los Angeles area, and the cohort included White, Hispanic, Black, and Asian women. The published analyses reported racial/ethnic breakdown but did not power for race-specific treatment effects.

In the early-postmenopause stratum, the direction of CIMT benefit with estradiol was consistent across racial groups. No group showed a signal in the opposite direction (harm). However, the cell sizes for non-White subgroups were small (generally fewer than 50 per treatment arm per race), making formal interaction testing unreliable.

This is an acknowledged limitation. The KEEPS trial, which studied a similar early-postmenopause population, had an even less diverse cohort. Neither trial can make definitive statements about racial differences in HRT cardiovascular response. The clinical takeaway is that no red flag emerged for any racial subgroup, but the evidence base for non-White women remains thin.

What the Subgroup Data Do Not Tell Us

Several important clinical questions remain unanswered by ELITE's subgroup analyses.

Genetic polymorphisms. Estrogen receptor alpha (ESR1) polymorphisms affect vascular estrogen signaling. ELITE did not genotype participants or analyze treatment response by ESR1 variant. Future trials incorporating pharmacogenomics could substantially refine patient selection.

Formulation effects. ELITE used oral 17-beta estradiol exclusively. Transdermal estradiol avoids first-pass hepatic metabolism and may have different cardiovascular effects, particularly on coagulation markers. The subgroup data cannot be extrapolated to patches, gels, or other delivery systems.

Hard cardiovascular endpoints. CIMT is a surrogate marker. ELITE was not powered for myocardial infarction, stroke, or cardiovascular death. The subgroup trends observed for CIMT may or may not translate to event reduction. The WHI trial remains the only large RCT with hard cardiovascular endpoints for HRT, and its population was predominantly late-postmenopause, making it a poor comparator for ELITE's early-initiation findings.

Progesterone interaction. Women with an intact uterus received vaginal progesterone in addition to estradiol. The published subgroup analyses did not separate hysterectomized women (estradiol alone) from those receiving combination therapy in a way that isolates progesterone's contribution to CIMT outcomes.

Translating Subgroups to Prescribing Decisions

Taken together, ELITE's subgroup data support a specific clinical profile for maximal cardiovascular benefit from early estradiol therapy.

The strongest signal of benefit was observed in women who were within 6 years of menopause onset, under age 55, with normal or mildly overweight BMI, and who had measurable (but early) subclinical atherosclerosis on CIMT. None of these factors operated independently of the timing stratum, which remains the dominant predictor.

For clinicians, this means the timing conversation should come first. A woman 3 years past menopause asking about HRT has a biological window where cardiovascular benefit is plausible. A woman 15 years past menopause does not, regardless of other favorable features. The 2022 Menopause Society guidelines reflect this hierarchy, placing time since menopause as the primary consideration before individual risk factors.

Frequently asked questions

›

References

Hodis HN, Mack WJ, Henderson VW, et al. Vascular effects of early versus late postmenopausal treatment with estradiol. N Engl J Med. 2016;374(13):1221-1231. PubMed
The NAMS 2017 Hormone Therapy Position Statement Advisory Panel. The 2017 hormone therapy position statement of The North American Menopause Society. Menopause. 2017;24(7):728-753. PubMed
Harman SM, Black DM, Naftolin F, et al. Arterial imaging outcomes and cardiovascular risk factors in recently menopausal women: a randomized trial (KEEPS). Ann Intern Med. 2014;161(4):249-260. PubMed
Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women's Health Initiative randomized controlled trial. JAMA. 2002;288(3):321-333. PubMed
FDA. Estradiol tablets prescribing information. Revised 2022. FDA Label
Polak JF, Pencina MJ, O'Leary DH, D'Agostino RB. Common carotid artery intima-media thickness progression as a predictor of stroke in MESA. Stroke. 2011;42(11):3017-3021. PubMed