What was the absolute risk increase for breast cancer in the WHI E+P trial?

The absolute increase was 8 additional invasive breast cancers per 10,000 women per year of use. Over the 5.2-year mean follow-up, this translated to roughly 4 extra cases per 1,000 treated women compared with placebo.

Why was the WHI E+P arm stopped early?

The Data and Safety Monitoring Board stopped the trial at the fifth interim analysis because the global index (a composite of CHD, stroke, PE, breast cancer, colorectal cancer, hip fracture, endometrial cancer, and death) crossed its predefined stopping boundary, indicating overall risks exceeded benefits.

Did the WHI test bioidentical hormones?

No. The WHI E+P arm used conjugated equine estrogens (Premarin) plus medroxyprogesterone acetate (Provera). It did not test transdermal estradiol, oral estradiol, or micronized progesterone. Results may not apply to other formulations.

What was the average age of women in the WHI?

Mean age at enrollment was 63.3 years. Only about one-third of participants were aged 50, 59, and the majority were more than 10 years past menopause onset at the time of randomization.

Did younger women in the WHI also have increased CHD risk?

In women aged 50, 59, the CHD hazard ratio was 1.29 (95% CI 0.79, 2.12), which was not statistically significant. The confidence interval includes both possible harm and possible benefit. The subgroup was too small to draw definitive conclusions.

What is the timing hypothesis for HRT?

The timing hypothesis proposes that estrogen is cardioprotective when started in early menopause (healthy arteries) but harmful when started late (atherosclerotic arteries). The ELITE trial provided supportive evidence. The WHI's older cohort was biased toward detecting harm rather than early-initiation benefit.

How did the WHI affect HRT prescribing in the United States?

U.S. prescriptions for combined estrogen-progestin products dropped approximately 46% within one year of the 2002 publication. The FDA added a black box warning, and clinical guidelines shifted from long-term prevention to short-duration symptom management.

Does the WHI apply to women taking HRT for menopausal symptoms?

Only 12.5% of WHI participants reported moderate-to-severe vasomotor symptoms at baseline. The trial population does not represent the typical symptomatic woman who seeks HRT. Current guidelines support use in symptomatic women under 60 with individualized risk assessment.

Was there a benefit for fractures in the WHI E+P arm?

Yes. Combined HRT reduced hip fractures by 34% (HR 0.66 to 95% CI 0.45, 0.98) and total fractures by 24% (HR 0.76 to 95% CI 0.69, 0.85). This benefit was consistent across all years of follow-up.

What was the venous thromboembolism risk with CEE + MPA?

Total VTE risk more than doubled (HR 2.11 to 95% CI 1.58, 2.82). Pulmonary embolism specifically increased by 113% (HR 2.13 to 95% CI 1.39, 3.25). The risk was highest in the first year and remained elevated throughout treatment.

WHI E+P Results in Detail: Numbers, Subgroups, and Time Course

By HealthRX Medical Team

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

At a glance

| Field | Detail | |---|---| | Trial | Women's Health Initiative Estrogen + Progestin (WHI E+P) | | N | 16,608 | | Intervention | Conjugated equine estrogens (CEE) 0.625 mg/d + medroxyprogesterone acetate (MPA) 2.5 mg/d | | Comparator | Matching placebo | | Duration | Mean 5.2 years (planned 8.5 years; stopped early) | | Primary endpoints | Coronary heart disease (nonfatal MI + CHD death) and invasive breast cancer | | Key result | CHD HR 1.29 (95% CI 1.02, 1.63); invasive breast cancer HR 1.26 (95% CI 1.00, 1.59) |

Why This Trial Exists

Before the WHI, combined hormone therapy was prescribed to millions of postmenopausal women based largely on observational data suggesting cardiovascular protection. The Nurses' Health Study and similar cohorts reported 40 to 50% reductions in CHD among hormone users. But observational studies carry selection bias: women who choose HRT tend to be healthier, leaner, and more engaged with preventive care. The WHI was designed as the definitive randomized test of whether CEE + MPA actually prevented heart disease and whether the suspected breast cancer signal was real. It enrolled 16,608 women aged 50, 79 across 40 U.S. clinical centers between 1993 and 1998.

Trial Design: What the Protocol Actually Specified

Participants were randomized 1:1 to one tablet daily of CEE 0.625 mg combined with MPA 2.5 mg or an identical placebo. The Data and Safety Monitoring Board (DSMB) used a modified Haybittle-Peto boundary with asymmetric stopping rules. The primary safety endpoint (invasive breast cancer) had a predefined boundary set at a weighted log-rank statistic of Z ≥ 3.0, adjusted for the multiple interim looks.

Two primary outcomes ran in parallel. CHD was the primary efficacy endpoint; invasive breast cancer was the primary safety endpoint. A "global index" combined these with stroke, pulmonary embolism, endometrial cancer, colorectal cancer, hip fracture, and death from other causes. If the global index indicated that overall risks exceeded benefits, the DSMB could recommend early termination. That is exactly what happened at the fifth interim analysis in May 2002.

Adherence declined over the study period. By year 6, only 54% of women randomized to active treatment were still taking study pills. In the placebo arm, approximately 10.7% initiated open-label hormone therapy during follow-up. These crossovers diluted the observed treatment effect, meaning the true biologic effect of CEE + MPA was likely larger than the intention-to-treat estimates captured.

Primary Endpoint Results

Coronary Heart Disease

The CEE + MPA group experienced 164 CHD events versus 122 in placebo over 5.2 years of mean follow-up. The hazard ratio was 1.29 (95% CI 1.02, 1.63), with an annualized rate of 0.37% in the hormone group and 0.30% in placebo. That translates to 7 additional CHD events per 10,000 person-years of use.

The time-course pattern was striking. CHD risk was highest in year 1 (HR 1.81 to 95% CI 1.09, 3.01) and attenuated over subsequent years. By years 4, 5, the point estimate moved toward the null. This early hazard spike aligns with prothrombotic and proinflammatory effects of oral estrogens observed in the HERS trial, which reported a similar first-year CHD increase in women with pre-existing heart disease.

Invasive Breast Cancer

The hormone group recorded 166 cases of invasive breast cancer compared with 124 in placebo. The hazard ratio was 1.26 (95% CI 1.00, 1.59). The annualized rate was 0.38% versus 0.30%, yielding 8 additional breast cancers per 10,000 person-years.

The breast cancer signal did not emerge immediately. In years 1, 2, hazard ratios were close to 1.0. The excess risk became apparent after year 3 and widened through year 5. Tumors in the CEE + MPA group were slightly larger at diagnosis (mean 1.7 cm vs. 1.5 cm) and more likely to be node-positive (25.9% vs. 15.8%), suggesting that combined HRT may have promoted growth of occult tumors rather than initiating new ones. This pattern has been confirmed in longer-term WHI follow-up data.

Secondary Endpoints: The Full Risk-Benefit Ledger

The table below shows all monitored outcomes from the original 2002 publication, expressed as hazard ratios with 95% confidence intervals and annualized absolute rates.

| Outcome | HR (95% CI) | CEE+MPA rate (%/yr) | Placebo rate (%/yr) | Excess per 10,000 PY | |---|---|---|---|---| | CHD (nonfatal MI + CHD death) | 1.29 (1.02, 1.63) | 0.37 | 0.30 | +7 | | Invasive breast cancer | 1.26 (1.00, 1.59) | 0.38 | 0.30 | +8 | | Stroke | 1.41 (1.07, 1.85) | 0.29 | 0.21 | +8 | | Pulmonary embolism | 2.13 (1.39, 3.25) | 0.17 | 0.08 | +9 | | Venous thromboembolism (total) | 2.11 (1.58, 2.82) | 0.34 | 0.16 | +18 | | Colorectal cancer | 0.63 (0.43, 0.92) | 0.10 | 0.16 | −6 | | Hip fracture | 0.66 (0.45, 0.98) | 0.10 | 0.15 | −5 | | Total fractures | 0.76 (0.69, 0.85) | 1.47 | 1.91 | −44 | | Endometrial cancer | 0.83 (0.49, 1.40) | 0.06 | 0.07 | −1 | | Death (all cause) | 0.98 (0.82, 1.18) | 0.52 | 0.53 | −1 | | Global index | 1.15 (1.03, 1.28) | 1.70 | 1.48 | +19 |

Source: Table 3 from Rossouw et al., JAMA 2002.

The global index, summing all monitored outcomes, showed a 15% increase in net harm. Pulmonary embolism showed the single largest relative increase (HR 2.13). Fracture protection and the colorectal cancer reduction partially offset the harms but were insufficient to tip the balance.

Subgroup Analyses by Age

The most consequential reanalysis of WHI data involved age stratification. Women aged 50, 59 at enrollment (n = 3 to 546 in the E+P arm) showed a different risk profile than women aged 70, 79.

| Age group | CHD HR (95% CI) | Breast cancer HR (95% CI) | Stroke HR (95% CI) | |---|---|---|---| | 50, 59 | 1.29 (0.79, 2.12) | 1.06 (0.71, 1.58) | 0.96 (0.49, 1.87) | | 60, 69 | 1.03 (0.73, 1.46) | 1.36 (1.00, 1.86) | 1.56 (1.05, 2.31) | | 70, 79 | 1.48 (1.04, 2.11) | 1.37 (0.89, 2.12) | 1.60 (1.00, 2.55) |

Data from the age-stratified WHI reanalysis.

For women in their 50s (within 10 years of menopause), CHD and stroke point estimates were closer to null, though confidence intervals were wide. The "timing hypothesis," formalized later by Hodis et al. in the ELITE trial, proposed that estrogen is cardioprotective when initiated in early menopause but harmful when started in women with established atherosclerosis. The WHI's mean enrollment age of 63 years meant most participants were well past menopause onset, a design choice that biased the trial toward detecting harm in an older population.

Time-Course Patterns Worth Knowing

The temporal behavior of risk differed by outcome.

Early hazards (year 1, 2): Venous thromboembolism and CHD both spiked in the first year. The VTE hazard ratio in year 1 exceeded 4.0 before settling to approximately 2.0 in later years. The oral route of estrogen delivery, which increases hepatic clotting factor synthesis, is likely responsible for this prothrombotic window. Transdermal estradiol bypasses first-pass hepatic metabolism. French and UK cohort data suggest that transdermal formulations carry substantially lower VTE risk.

Delayed hazards (year 3+): Breast cancer excess appeared only after the third year of continuous use. This latency is consistent with a promoter rather than initiator mechanism and matches patterns in the Million Women Study observational data.

Sustained benefits: Fracture reduction was consistent across all follow-up years. The colorectal cancer benefit also persisted, though it attenuated in extended post-intervention follow-up.

Limitations the Authors Acknowledged

The original publication listed several constraints that are routinely overlooked when the trial is cited.

Single formulation tested. CEE 0.625 mg + MPA 2.5 mg was the only regimen studied. The WHI results cannot be directly extrapolated to bioidentical estradiol, micronized progesterone, lower doses, or transdermal delivery. The 2022 Menopause Society position statement notes that micronized progesterone may carry a different breast cancer risk profile than MPA.

Older population. Mean age at enrollment was 63.3 years. Only 33% of participants were aged 50, 59. The trial was not designed to test the timing hypothesis and was underpowered to detect benefit in recently menopausal women.

Adherence erosion. With 42% of the active group off study drug by the end and over 10% of the placebo group on open-label hormones, the ITT analysis underestimates biological effect in both directions (harm and benefit).

Ethnicity. The cohort was 84% non-Hispanic White, limiting generalizability.

No symptom-driven enrollment. Only 12.5% of participants reported moderate-to-severe vasomotor symptoms at baseline. The trial population does not reflect the typical clinical scenario where HRT is prescribed to manage hot flashes and sleep disruption.

What Changed After Publication

Within 12 months of the July 2002 publication, U.S. prescriptions for Prempro dropped by 46%. Millions of symptomatic women discontinued therapy, sometimes abruptly. The FDA mandated a black box warning on all systemic estrogen-progestin products referencing the WHI findings. Professional society guidelines shifted from recommending HRT for chronic disease prevention to restricting it to symptom management at the lowest effective dose for the shortest necessary duration.

Two decades of subsequent reanalysis have moderated the initial alarm. The age-stratified data, the timing hypothesis, the distinction between oral CEE + MPA and other formulations, and the recognition that observational selection bias may not explain all of the pre-WHI benefit signals have led to a more nuanced clinical framework. Current Endocrine Society guidelines support initiating HRT in symptomatic women under 60 or within 10 years of menopause, with individualized risk-benefit assessment.

Frequently asked questions

›

References

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
Rossouw JE, Prentice RL, Manson JE, et al. Postmenopausal hormone therapy and risk of cardiovascular disease by age and years since menopause. JAMA. 2007;297(13):1465-1477. 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
Chlebowski RT, Anderson GL, Gass M, et al. Estrogen plus progestin and breast cancer incidence and mortality in postmenopausal women. JAMA. 2010;304(15):1684-1692. PubMed
Premarin/MPA prescribing information. U.S. Food and Drug Administration. FDA Label
The 2022 Hormone Therapy Position Statement of The North American Menopause Society. Menopause. 2022;29(7):767-794. PubMed