Did KEEPS prove the timing hypothesis?

Not definitively. The trial showed that early HRT did not increase atherosclerosis markers, which is consistent with the timing hypothesis. But the primary CIMT endpoint was null, so KEEPS provides supportive rather than confirmatory evidence. ELITE's positive CIMT finding in the early-initiation group strengthens the case.

Is oral CEE or transdermal estradiol better based on KEEPS?

Neither formulation beat placebo on the primary endpoint. Both improved symptoms comparably. Transdermal estradiol avoids hepatic first-pass metabolism, which has implications for clotting risk, triglycerides, and SHBG. Most menopause specialists now prefer transdermal delivery for patients with metabolic or thrombotic risk factors.

Why did KEEPS use micronized progesterone instead of medroxyprogesterone?

The investigators chose micronized progesterone based on emerging evidence that it carries lower breast cancer and cardiovascular risk than MPA. This design choice has since been validated by observational studies and endorsed by NAMS and the Endocrine Society.

Can KEEPS results be applied to women with diabetes or obesity?

No. KEEPS excluded women with BMI ≥35 and those on diabetes or antihypertensive medications. Applying KEEPS safety data to metabolically unhealthy patients requires extrapolation the trial does not support.

Did KEEPS find any cardiovascular benefit at all?

The oral CEE arm showed a non-significant trend toward less CIMT progression and lower coronary calcium scores. These observations are hypothesis-generating but did not reach statistical significance.

How long should a woman stay on HRT based on KEEPS?

KEEPS only covered four years. It cannot guide duration-of-use decisions. Current NAMS guidance recommends reassessing annually and using the lowest effective dose for the shortest needed duration, though it no longer imposes arbitrary time limits.

Should HRT be prescribed specifically to prevent heart disease?

No. No current guideline recommends HRT for primary cardiovascular prevention. KEEPS reinforces that early HRT is not cardiotoxic, but that is different from being cardioprotective. Prescribing HRT for symptom management in appropriate candidates remains the evidence-based indication.

What is the difference between KEEPS and WHI?

Population age (early vs. late menopause), dose (low vs. standard), progestogen type (micronized progesterone vs. MPA), and endpoint (surrogate imaging vs. clinical events). These differences are large enough that the two trials answer fundamentally different clinical questions.

Did KEEPS change FDA labeling for estrogen products?

Not directly. FDA labeling for estrogen products still carries the class-wide WHI-derived boxed warning. However, the 2017 FDA guidance acknowledged that WHI risks may not apply equally to younger, recently postmenopausal women, a position KEEPS data helped support.

Are there ongoing follow-up studies from KEEPS participants?

KEEPS-Continuation followed a subset of participants after the trial ended, examining cognitive outcomes and sustained vascular effects. Early results suggested no lasting cognitive harm, but long-term cardiovascular event data from the original cohort remains limited.

What KEEPS Actually Changes in Clinical Practice

By HealthRX Medical Team

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

At a glance

| Field | Detail | |---|---| | Trial Name | Kronos Early Estrogen Prevention Study (KEEPS) | | N | 727 recently postmenopausal women | | Intervention | Oral conjugated equine estrogen (o-CEE) 0.45 mg/day OR transdermal 17β-estradiol (t-E2) 50 mcg/day, both with cyclic oral micronized progesterone 200 mg × 12 days/month | | Comparator | Matching placebo | | Duration | 48 months | | Primary Endpoint | Rate of change in carotid artery intima-media thickness (CIMT) | | Key Result | Neither HRT formulation significantly slowed CIMT progression vs. placebo; both improved vasomotor symptoms, mood, bone density, and sexual function | | Published | 2014, Annals of Internal Medicine |

Why KEEPS Existed at All

The 2002 Women's Health Initiative (WHI) results triggered a near-total retreat from postmenopausal hormone therapy. Prescriptions dropped by roughly 80% within five years. But the WHI enrolled women whose average age was 63, most of them more than a decade past menopause. Critics argued that starting HRT late, in women who already had subclinical atherosclerosis, was a fundamentally different intervention than starting it during the menopausal transition.

That argument became the "timing hypothesis." KEEPS was designed to test it directly: enroll women aged 42 to 58 who were within 36 months of their final menstrual period, give them low-dose estrogen (lower than the WHI doses), and measure whether early initiation slowed arterial aging as tracked by CIMT.

Methodology Worth Reading Closely

Several design choices set KEEPS apart from WHI and deserve attention when translating results to practice.

Dose selection. The oral CEE arm used 0.45 mg/day, half the WHI dose. The transdermal arm used a 50 mcg/day patch. Both are doses commonly prescribed today for symptom management, which makes the safety data directly applicable to current practice rather than an artifact of higher, less common dosing.

Progesterone choice. KEEPS used oral micronized progesterone (Prometrium), not medroxyprogesterone acetate (MPA, the progestin used in WHI). The 2017 Endocrine Society guideline later cited accumulating evidence, including KEEPS, that micronized progesterone carries a better risk profile for breast and cardiovascular endpoints than synthetic progestins.

Population strictness. Enrolled women had to be within 6 to 36 months of their last period, with a BMI <35, no diabetes, no prior cardiovascular events, and no statin or antihypertensive use at screening. This created a uniformly healthy, early-postmenopausal cohort. The advantage: a clean test of the timing hypothesis. The limitation: results say little about women with existing metabolic disease.

Imaging endpoints. The primary outcome was annual CIMT change. A secondary endpoint, coronary artery calcium (CAC) scoring by CT, was measured at baseline and 48 months. CIMT is a validated surrogate for atherosclerosis progression but is not a clinical event endpoint. No version of KEEPS was powered to detect differences in heart attacks or strokes.

Results: What the Numbers Actually Show

CIMT (Primary Endpoint)

| Group | Annual CIMT Change (mm/yr) | Difference vs. Placebo | |---|---|---| | Placebo | 0.007 |, | | Oral CEE 0.45 mg | 0.005 | −0.002 (95% CI: −0.006 to 0..002, p = 0.29) | | Transdermal E2 50 mcg | 0.007 | 0.000 (95% CI: −0.004 to 0.003, p = 0.98) |

Neither arm reached statistical significance for slowing CIMT progression. The oral CEE arm trended toward less progression, but the confidence interval crossed zero. The transdermal arm showed no trend at all. These null primary results were disappointing to timing-hypothesis proponents expecting a clear vascular benefit.

Coronary Artery Calcium (Secondary)

CAC scores at 48 months were lower in the oral CEE group than placebo (median Agatston score difference favoring CEE), but this difference was not significant after adjustment. The transdermal arm showed no CAC advantage. CAC was a secondary endpoint with limited power, so these results are hypothesis-generating only.

Symptom and Quality-of-Life Outcomes

Both HRT arms significantly improved:

Hot flash frequency and severity (p <0.001 vs. placebo for both formulations)
Mood and depressive symptoms (measured by PHQ-9 and CES-D scales)
Sexual function scores
Bone mineral density at spine and hip

Sleep quality improved more with transdermal estradiol than oral CEE, a finding consistent with the lower hepatic first-pass effect of patch delivery.

Safety Signals

Over four years, KEEPS recorded no increase in breast cancer, stroke, or venous thromboembolism in either HRT arm versus placebo. The study was not powered to detect rare events, so absence of signal is not proof of absence. But the data was reassuring enough to influence subsequent guideline language about early-initiation safety.

The HealthRX Clinical Translation Framework

We use a four-question filter when deciding what a trial should actually change in the clinic, rather than just what it proved statistically.

1. Did the result change a guideline? Yes. The 2017 North American Menopause Society (NAMS) position statement cited KEEPS when affirming that for women under 60 or within 10 years of menopause, the benefits of HRT for vasomotor symptoms "generally outweigh the risks." The Endocrine Society's 2017 guideline similarly referenced KEEPS data when recommending transdermal estradiol as a preferred option for women with elevated thrombotic risk. Before KEEPS, many clinicians treated WHI as a blanket prohibition. After KEEPS, the messaging became: timing, dose, and formulation selection matter.

2. Did prescribing patterns actually shift? Partially. National prescription data from IQVIA show that transdermal estradiol prescriptions increased as a proportion of total HRT scripts from 2014 onward, and micronized progesterone gained market share against MPA. These shifts align with the KEEPS protocol. Absolute HRT prescribing rates remain well below pre-WHI levels, suggesting persistent physician and patient anxiety that KEEPS alone could not reverse.

3. Does the result apply to patients who differ from the trial population? This is where caution matters most. KEEPS excluded women with BMI ≥35, diabetes, hypertension on medication, and prior cardiovascular events. A 52-year-old patient with metabolic syndrome asking about HRT is not a KEEPS patient. Extrapolating KEEPS safety data to her carries real risk. Similarly, KEEPS enrolled predominantly white women (over 90%), limiting generalizability to other racial and ethnic groups where cardiovascular risk profiles and estrogen metabolism may differ.

4. What question does the trial leave unanswered? KEEPS ran for four years. It cannot tell clinicians what happens at year eight or year twelve. The ongoing ELITE trial (published 2016) provided complementary data showing that early-initiation estradiol did slow CIMT progression over five years, but only in younger women, reinforcing the timing hypothesis from a different angle. Neither trial answers the hard question: does early HRT initiation reduce actual cardiovascular events over a decade or more? That trial has never been done, and the ethical and logistical barriers to running it are substantial.

What Should Change in Practice Today

Start the conversation earlier. KEEPS supports initiating the HRT discussion during the menopausal transition, not years later when symptoms have already degraded quality of life and the vascular window may have closed. The NAMS "10-year window" recommendation draws partly from this trial.

Default to low-dose formulations. The 0.45 mg oral CEE and 50 mcg transdermal patch used in KEEPS managed symptoms effectively. There is limited justification for starting at higher doses in most patients.

Prefer transdermal delivery for thrombotic risk. KEEPS showed no VTE signal in either arm, but the trial was underpowered for this outcome. Observational data from the ESTHER study and others consistently show lower clotting risk with transdermal versus oral estrogen. For patients with obesity, migraine with aura, or family history of VTE, transdermal is the defensible default.

Use micronized progesterone, not MPA. KEEPS protocol used Prometrium. Combined with the E3N cohort data showing lower breast cancer risk with micronized progesterone than with synthetic progestins, this is now standard in most menopause-focused practices.

Do not promise cardiovascular protection. KEEPS failed its primary endpoint. The trial supports the claim that early HRT does not accelerate atherosclerosis. It does not support prescribing HRT for cardiovascular prevention. The distinction matters for informed consent. A clinician telling a patient "this will protect your heart" is overstating the evidence. "This is unlikely to harm your heart if started early" is the accurate framing.

Limitations the Authors Acknowledged

The original investigators were transparent about several constraints. The sample size (n=727) was powered for CIMT, a continuous imaging variable, not clinical events. CIMT itself is an imperfect surrogate; its correlation with cardiovascular events has been questioned in meta-analyses. The placebo group showed very slow CIMT progression (0.007 mm/year), creating a low baseline rate that was difficult to improve upon. Dropout was approximately 15%, which reduced statistical power further. And the four-year duration may have been too short to detect a separation between treatment and placebo in a healthy population aging slowly.

How KEEPS Compares to Adjacent Trials

| Trial | Population | Duration | Primary Endpoint | Result | |---|---|---|---|---| | KEEPS | Early postmenopausal (<3 yr), healthy | 4 yr | CIMT | Null for CIMT; symptom benefit | | ELITE | Early (<6 yr) vs. late (≥10 yr) postmenopausal | 5 yr | CIMT | Benefit in early group only | | WHI | Late postmenopausal (mean age 63) | 5.6 yr | CV events + cancer | Increased stroke and VTE; no CHD benefit | | DOPS | Early postmenopausal | 10 yr | Composite CV + death | 52% reduction in primary composite |

The Danish Osteoporosis Prevention Study (DOPS) remains the only long-term RCT showing a cardiovascular event reduction with early HRT. Its open-label design limits the strength of that finding, but the 10-year duration provides data KEEPS and ELITE cannot.

Frequently asked questions

›

References

Harman SM, Black DM, Naftolin F, et al. Arterial imaging outcomes and cardiovascular risk factors in recently menopausal women: a randomized trial. Ann Intern Med. 2014;161(4):249-260. PubMed
Hodis HN, Mack WJ, Henderson VW, et al. Vascular effects of early versus late postmenopausal treatment with estradiol (ELITE). 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
Schierbeck LL, Rejnmark L, Tofteng CL, et al. Effect of hormone replacement therapy on cardiovascular events in recently postmenopausal women (DOPS). BMJ. 2012;345:e6409. PubMed
Canonico M, Oger E, Plu-Bureau G, et al. Hormone therapy and venous thromboembolism among postmenopausal women: impact of the route of estrogen administration and progestogens (ESTHER study). Circulation. 2007;115(7):840-845. PubMed
Stuenkel CA, Davis SR, Gompel A, et al. Treatment of symptoms of the menopause: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2015;100(11):3975-4011. PubMed