More from the Women’s Health Initiative

A Special Report by Leon Speroff

The latest news from the women’s health initiative (WHI) includes 3 noteworthy reports: the news release announcing the cancellation of the estrogen-only arm of the clinical trial,1 a comparison of the participants in the 2 clinical trial arms (estrogen-progestin and estrogen-only),2 and the updated, adjudicated colorectal cancer results from the estrogen-progestin arm.3

The Estrogen-only Arm of the Clinical Trial

On March 2, 2004, the National Heart, Lung, and Blood Institute of the US National Institutes of Health (NIH) canceled the estrogen-only (0.625 mg conjugated estrogens daily) arm of the WHI multicenter, randomized, clinical trial. This arm of the WHI included almost 11,000 hysterectomized, postmenopausal women who had completed an average of nearly 7 years of follow-up. The WHI Data and Safety Monitoring Board (DSMB) made its last periodic review of the study data in December 2003. The DSMB was not unanimous in its decision; some wanted to stop the study and others wanted the study to continue after sending a letter to the participants describing the findings. The NIH made the decision to stop the study on February 2, 2004.

Why was the study stopped? The decision was based on the following results according to the NIH:

  • An increased risk of stroke similar to that reported in the canceled estrogen-progestin arm of the WHI;
  • No increase or decrease in coronary heart disease;
  • A trend toward an increased risk of probable dementia and/or mild cognitive impairment;
  • A reduction in hip fractures;
  • No increase in breast cancer.

The academic grapevine has indicated that a major consideration in making the decision to cancel the estrogen-only arm of the study was the high drop-out rate in response to the publicity of the last 2 years, severely eroding the statistical power and leaving nothing to be gained by continuing until the scheduled termination in 2005.

With the exception of breast cancer, the results are essentially identical to those in the estrogen-progestin arm of the study. Indeed, according to the grapevine, the results with breast cancer in the estrogen-only arm were actually indicating a statistically significant decrease in risk. But keep in mind, that the increase in breast cancer in the estrogen-progestin arm was a small one, that the statistical power of the estrogen-only arm is not as great (6000 fewer participants), and the populations of the 2 arms may not be comparable (especially different ages of menopause). Indeed, the populations in the 2 clinical, randomized trial arms of the WHI were not identical.2 Considering risk factors for cardiovascular disease, the women in the estrogen-only arm were more obese, less active, and had more preexisting cardiovascular disease. The estrogen-only arm also differed in regard to risk factors for breast cancer: more early births, bilateral oophorectomy, and more and longer duration of previous hormone therapy. In the coming months, it will be tempting to compare the 2 canceled arms of the WHI; however, these were 2 different trials with 2 different populations and treatments, making direct comparisons inappropriate.

Clinicians and patients should regard the cancellation of the estrogen-only arm of the clinical trial as good news. Clinicians and patients should accept this new report as another indicator that postmenopausal hormone therapy does not have a major effect on the risk of breast cancer. When the published results become available, it will be important to carefully analyze the statistics. The estrogen-only arm may lack the power to make a confident statement regarding breast cancer for 2 reasons: the estrogen-only arm is about 62% of the size of the estrogen-progestin arm, and a high drop-out rate may have eroded its statistical strength. In addition, as noted earlier, important risk factors that favor a reduced risk for breast cancer were predominant in the estrogen-only arm.

Like the estrogen-progestin arm, it is reasonable to expect the stroke and dementia/cognition results to reflect the effect of this dose of estrogen given to an older group of women, many years distant from menopause. According to the NIH news release, the average age of the participants in the estrogen-only arm is now nearly 70, indicating that the ages in the 2 arms of the WHI clinical trial were comparable. Most importantly, the published comparison of the 2 clinical arms indicated that preexisting cardiovascular disease was even more prevalent in the estrogen-only arm.2 In the canceled estrogen-progestin arm, the only detrimental effect was an increase in dementia in the group of women who were 75 years and older when they started treatment, and the dementia was probably vascular in origin, not Alzheimer’s. Because the ages in the 2 arms are comparable and there was more preexisting cardiovascular disease in the estrogen-only arm, it is likely that the dementia results are the same, an effect only in very old women who already have significant atherosclerosis.

The report of neither a beneficial effect nor an adverse effect on coronary heart disease gives the impression that the estrogen-only arm represents an improvement over the estrogen-progestin arm. The updated results on the risk of coronary heart disease from the canceled estrogen-progestin arm of the WHI reflect central adjudication of the cardiac diagnoses in contrast to the initial report that relied on local diagnoses.4 Local and central reviews disagreed with 10% of the diagnoses for myocardial infarction and 3% for death due to coronary heart disease. This small degree of disagreement changed the strength of the conclusions comparing the initial report5 with the updated report. Indeed, the overall results by definition did not achieve statistical significance in the follow-up report, and only the first-year results were statistically significant in the year-by-year analysis, a conclusion based on a difference of only 19 cases. In the subgroup analyses, only the women who were 20 or more years distant from menopause had a statistically significant increased risk of coronary heart disease (1.71; CI = 1.20-2.50). Subtracting this group from the rest of the participants, coronary heart disease now was observed in an identical prevalence comparing the treated and placebo groups. It is not appropriate to conclude, based on the WHI, that hormone therapy increases the risk of coronary clinical events in all postmenopausal women; this conclusion can only be applied to a specific older group of women. The WHI did not study a young postmenopausal, healthy population in order to establish that hormone therapy does not exert a primary preventive effect on the risk of coronary heart disease. That is not to say that the case for primary prevention of coronary heart disease by postmenopausal hormone therapy does not merit controversy. However, at the same time, the issue is not settled, and perhaps there never will be a large trial of the appropriate population. We cannot quarrel with the conclusion that postmenopausal hormone therapy does not reduce or slow the progression of established coronary heart disease. However, the WHI did not study the appropriate population in the appropriate time period to establish that hormone therapy does not exert a primary preventive effect on the risk of coronary heart disease.

Colorectal Cancer

Most, but not all, cohort and case-control studies have reported a significantly reduced risk of colorectal cancer incidence and mortality in users of postmenopausal estrogen.6-12 The effect is greatest in current users and most studies have not indicated an increased effect with increasing duration of use; for example, the Nurses’ Health Study (which found a 34% reduced risk in current users) could not demonstrate an added benefit with longer duration of current use.13 A reduction in fatal colon cancer has been documented in current users.8 There also appears to be a reduced risk of polyps, especially large polyps, among current and recent hormone users.

The canceled estrogen-progestin arm of the WHI reported a statistically significant reduced risk of colon cancer achieved with only a few years of estrogen-progestin therapy.3 There were too few cases of rectal cancer to allow separate analysis (see Table below).

This result was not without concern, however, in that the treated group had more advanced disease. Indeed, the conclusion was largely because of a difference in localized disease, 10 cases in the treated group and 36 in the placebo group. The results suggest that already present cancers were influenced by hormone therapy to reach a more advanced stage, but that estrogen-progestin treatment reduced the risk of new colonic cancers.

One can only speculate regarding the mechanism of this benefit. The estrogen-induced change in the bile (a decrease in bile acids with an increase in cholesterol saturation) favors gallstone formation but may reduce promotion (by bile acids) of colonic cancer. Other possible mechanisms include a direct suppressive effect on mucosal cell growth and an effect on beneficial mucosal secretions. The colon contains only estrogen receptor-beta, and the reduction in the risk of colonic cancer associated with postmenopausal estrogen therapy may reflect an antiproliferative activity of the beta receptor. This potential benefit deserves greater attention; colorectal cancer ranks third in women, both in incidence and mortality, and is more prevalent than cancers of the uterus or ovary.14


So the clinical arms of the WHI are over. But, of course, we will continue to hear from the WHI for many months as the reports periodically appear. I am confident that academicians and clinicians are now better prepared to apply a critical analysis to all of the WHI data. This is a process that takes time for the consideration and appraisal of the published data. For this reason, it doesn’t pay to give serious attention to those quoted experts that appear in the media during the day and week that research reports appear. I believed that, over time, the results of the WHI will be placed in their appropriate role in clinical decision making. Overall, the WHI results are not as negative (adverse effects) as many, especially the media, have portrayed.


1. National Heart, Lung, and Blood Institute. 2004.
2. Stefanick ML, et al. Ann Epidemiol. 2003;13:S78-S86.
3. Chlebowski RT, et al. N Engl J Med. 2004;350: 991-1004.
4. Manson JE, et al. N Eng J Med. 2003;349:523-534.
5. Writing Group for the Women’s Health Initiative Investigators. JAMA. 2002;288:321-333.
6. Chute CG, et al. Epidemiology. 1991;2:201-207.
7. Jacobs EJ, et al. Cancer Causes Control. 1994;5: 359-365.
8. Calle EE, et al. J Natl Cancer Inst. 1995;87:517-523.
9. Kampman E, et al. Cancer Causes Control. 1997;8: 146-158.
10. Troisi R, et al. Cancer Causes Control. 1997;8:130-138.
11. Fernandez E, et al. Cancer Epidemiol Biomarkers Prev. 1998;7:329-333.
12. Paganini-Hill A. Dis Colon Rectum. 1999;42: 1300-1305.
13. Grodstein F, et al. Ann Intern Med. 1998;128:705-712.
14. American Cancer Society. Cancer Facts & Figures 2003. 2003.

Leon Speroff, MD, Professor of Obstetrics and Gynecology, Oregon Health Sciences University, Portland, is Editor of OB/GYN Clinical Alert.