Discrepancies between reports of HRT dangers and supporting data

Apr 23 2009

In recent years, a series of highly publicized reports have warned of increases in breast cancer and other health problems in postmenopausal women who take hormone replacement therapy (HRT).

But a closer look at the data suggests that those reports were often "distorted, oversimplified, or wrong," according to a paper in the March/April issue of The Cancer Journal: The Journal of Principles & Practice of Oncology. The journal is published by Lippincott Williams & Wilkins, a part of Wolters Kluwer Health, a leading provider of information and business intelligence for students, professionals, and institutions in medicine, nursing, allied health, pharmacy and the pharmaceutical industry.

Despite "weak or statistically insignificant" data, scientific and media reports have emphasized the dangers of HRT, while downplaying its beneficial effects especially on menopausal symptoms, according to the article by Avrum Z. Bluming, M.D., and Carol Tavris, Ph.D. "Physicians and the public must be cautious about accepting 'findings by press release' in determining whether or not to prescribe or take HRT," according to Drs. Bluming and Tavris. Dr. Bluming is a Board-Certified Medical Oncologist, a Master of the American College of Physicians and a Clinical Professor of Medicine at the University of Southern California. Dr. Tavris is a social psychologist and writer.

Discrepancies Between Reports of HRT Dangers and Supporting Data
Doctors Bluming and Tavris cite the Harvard Nurses' Health Study, published in 1995, which employed retrospective substratification to exaggerate the association between HRT and breast cancer development. Statisticians have cautioned that spurious associations can result from this practice-also called "data mining"-in which researchers sift through their databases, looking for factors that might affect health outcomes. "The problem is that in a data set of many thousands of people, some relationship that is unearthed retrospectively will turn out to be statistically significant...just by chance."

The authors also criticize the Women's Health Initiative (WHI), published initially in 2002, which reported "relative risk" rather than "absolute risk" associations. This format can make "statistically modest or borderline results...look more impressive than they actually are." The authors note that the 26 percent relative risk of breast cancer in the initial WHI is similar to that reported for unlikely risk factors like eating grapefruit, working on a nightshift, or working as a flight attendant.

The core issue, according to Drs. Bluming and Tavris, is the difference between information and knowledge. "What we're trying to do is make a plea for knowledge, not just information, especially loosely based associations that are generated by epidemiologic studies. That isn't science. Post-hoc findings that make little or no scientific sense may give us clues for future investigation, but they shouldn't be accepted as proof of cause and effect."

Jumble of 'Positive, Negative and Meaningless' Findings
Press releases linking HRT to breast cancer failed to report that the initial WHI report of a 26 percent relative risk did not reach the level of statistical significance. Nevertheless, several editorials by prominent authorities highlighted the increase in risk as "statistically significant"-a false claim repeated in news stories.

Subsequent analyses gave varying results: smaller but significant risk, larger but nonsignificant risk, or no significant difference in risk. "It is difficult to resist the conclusion that the WHI investigators have been doing everything they could to wring the bleakest possible interpretation from their recalcitrant data," Drs. Bluming and Tavris write.

Although concerns about breast cancer and other health risks are valid, "HRT is not the clear and present danger that the WHI and much of the media have made it out to be," Drs. Bluming and Tavris conclude. They contend that the weight of available evidence supports the use of HRT by women who are having menopausal symptoms, beginning at the start of menopause and continuing for as many years as necessary.